English Pronunciation for Speakers of Spanish: From Theory to Practice [1° ed.] 1501510967, 9781501510960

Citation preview

List of Tables Table 1 Table 2 Table 3 Table 4 Table 5 Table 6 Table 7 Table 8 Table 9 Table 10 Table 11 Table 12 Table 13 Table 14 Table 15 Table 16 Table 17 Table 18 Table 19 Table 20 Table 21 Table 22 Table 23 Table 24 Table 25 Table 26 Table 27 Table 28 Table 29 Table 30 Table 31 Table 32 Table 33 Table 34 Table 35 Table 36

The interface of the auditory, acoustic and articulatory 10 levels 21 Syllable patterns in English and Spanish 32 The symbols of the IPA 37 The sound systems of RP and PSp 61 Primary cardinal vowels 65 Voiced and voiceless consonants in RP and PSp 75 Formant frequencies of RP vowels 88 Classification of the vowels of RP and PSp 93 The spellings of RP /i: ɪ/ 100 The spellings of RP /e ɜː ə/ 106 The spellings of RP /ʌ æ ɑː/ 113 The spellings of RP /ɔː ɒ/ 118 The spellings of RP /u: ʊ/ 127 The spellings of RP /eɪ aɪ ɔɪ/ 132 The spellings of RP /əʊ aʊ/ 136 The spellings of RP /ɪə eə ʊə/ 141 The spellings of RP diphthongs + [ə] 153 The distinctive consonants of RP and PSp Differences between the consonantal phonemic systems of English 155 and Spanish 167 The spellings of RP /p b/ 172 The spellings of RP /t d/ 179 The spellings of RP /k g/ 186 The spellings of RP /f v/ 189 The spellings of RP /θ ð/ 193 The spellings of RP /s z/ 200 The spellings of RP /ʃ ʒ/ 203 The spellings of RP /h/ 207 The spellings of RP /tʃ dʒ/ 213 The spellings of RP /m n ŋ/ 219 The spellings of RP /l r/ 225 The spellings of RP /j w/ 235 Main allophonic variations of vowels and consonants 245 Regressive assimilation 246 Progressive assimilation 246 Regressive nasalisations 246 Regressive assimilations: Voicing

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:53 PM

xii Table 37 Table 38 Table 39 Table 40 Table 41 Table 42 Table 43 Table 44 Table 45 Table 46 Table 47 Table 48 Table 49 Table 50 Table 51 Table 52 Table 53 Table 54 Table 55 Table 56 Table 57 Table 58 Table 59 Table 60 Table 61 Table 62 Table 63 Table 64 Table 65 Table 66 Table 67 Table 68

List of Tables

Progressive assimilations: Voicing 247 247 Coalescent assimilation 252 Weak and strong forms 256 Neutralisation of weak forms 257 Neutralisation matrix General tendencies in stress patterns: 2 and 3 syllable 266 words 267 Most common pairs of words with variable stress 279 Proposed intonation marking system 284 High tones 285 Rise tones 287 Complex tones: Fall-rises and rise-falls Accentuation disparities between English and Spanish 311 cognates 323 Vowels in closed and open syllables 323 Orthographic syllable division 324 Tense vowels in non-final syllables 324 Lax vowels in non-final syllables 325 Heavy and r-tense vowels 326 Vowels of the four groups 327 Sounds after RP /w/ 329 Vowels of the four groups (extended) 330 V letter followed by other V letter or 331 V followed by (V)C(C) 332 Digraphs + 337 Consonants: Text-to-speech correspondences 341 Lax vowels: Sound-to-spelling correspondences 342 Tense vowels: Sound-to-spelling correspondences 343 Heavy vowels: Sound-to-spelling correspondences 344 R-tense vowels: Sound-to-spelling correspondences 344 Other diphthongs: Sound-to-spelling correspondences 345 Schwa /ə/: Sound-to-spelling correspondences 345 Unstresed /ɪ/: Sound-to-spelling correspondences 346 Consonants: Sound-to-spelling correspondences

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:53 PM

List of Figures Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18

Figure 19 Figure Figure Figure Figure Figure Figure Figure Figure Figure

20 21 22 23 24 25 26 27 28

Figure Figure Figure Figure Figure Figure Figure

29 30 31 32 33 34 35

The speech chain 3 6 Periodic and aperiodic sinewaves 7 Sinewaves with different amplitudes and frequencies 7 Constructing a complex wave 8 Speech waveform and spectrogram of “passed London” 9 Narrowband and wideband spectrograms 15 Realisation rule for some allophones of /t/ 20 Plotting syllable prominence and syllable boundaries 26 The structure of the syllable in English 29 Sonority scale or sonority hierarchy 44 Systems involved in the production of speech 46 The breathing cycle 48 Larynx in cross section 48 Larynx viewed from above 49 Human larynx and vocal folds 50 Phonation modes 53 The articulatory system Highest tongue positions for (a) cardinal front vowels and (b) back 59 vowels Relative highest and lowest points of the tongue for cardinal 59 vowels 60 The Cardinal Vowel Scale 60 Primary cardinal vowels vs. secondary cardinal vowels 66 Seven places of articulation relevant for English 67 Places of articulation for RP consonants 68 Manner of articulation for plosive stops 69 Manner of articulation for nasal stops 70 Palatograms of fricative articulations 74 Reading a spectrogram Waveforms and spectrograms of “leakage”, “leak”, “lead” and 75 “lee” 76 Waveforms and spectrograms of “bee” and “pea” 77 Spectrograms of [aɪ eɪ ɔɪ aʊ əʊ eə ɪə ʊə] 78 Acoustic phases of plosives in [ɑpʰɑː] and [ɑbɑː] 78 Spectrograms showing the noise patterns of fricatives 79 Spectrograms of “ages” [ˈeɪʤɪz] and “h’s” [ˈeɪʧɪz] 80 Spectrograms of “sim” [sɪm], “sin” [sɪn] and “sing” [sɪŋ] 81 Spectrograms of [aˈja], [aˈwa], [aˈɹa], [aˈla]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:56 PM

xiv Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure Figure

List of Figures

36 37 38 39 40 41 42 43 44 45 46 47 48 49 50

The vowels of RP and PSp in the CVS 91 92 RP /iː ɪ/ vs. PSp /i/ in the CVS 98 RP /e ɜː ə/ vs. PSp /e/ 105 RP /ʌ æ ɑː/ vs. PSp /a/ 112 RP /ɒ ɔː/ vs. PSp /o/ 116 RP /u: ʊ/ vs. PSp /u/ 124 RP diphthongs: (a) closing, (b) centring 125 PSp diphthongs: (a) falling, (b) rising 126 [ɪ/i] diphthongs in RP and PSp 131 [ʊ/u] diphthongs in (a) RP and (b) PSp 135 Centring diphthongs in RP 140 Diphthongs + [ə] in RP 163 VOT values for RP plosives Examples of illustrating the VOT values for RP plosives Whatsapp Emoticons as a graphic representation of 290 attitude

163

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:56 PM

Audio Tracks All the listening material accompanying this book, whether audio exercises or audio illustrations, is available in the companion website of this book, EPSS Multimedia Lab (http://www.usc.es/multimlab/index.html). You can visit the Lab for your own choice of audio exercises as you proceed through the material in the book. Audio illustrations (listed below) are marked and a reference number, where relevant throughout the book with the icon AI 1.3 means ‘Audio Illustration 3 in Chapter 1’. You can easily find so that and listen to them in the EPSS Multimedia Lab under the “Audio Illustrations” tab in the main menu. Audio Illustrations Chapter 1 13 AI 1.1 Identification of phonemes 15 AI 1.2 Contrastive distribution of phonemes in RP AI 1.3 Identification of a short vowel as syllabic nucleus 29 AI 1.4 Syllabic consonants in RP 33 AI 1.5 Becoming familiar with the IPA 34 AI 1.6 An example of broad transcription 36 AI 1.7 The sound systems of RP and PSp

20

Audio Illustrations Chapter 2 45 AI 2.1 Ingressive and egressive airstream 51 AI 2.2 Glottal stop 51 AI 2.3 Creaky voice 57 AI 2.4 Vowels and glides of English RP 64 AI 2.5 Voiced and voiceless consonants 68 AI 2.6 RP plosives 69 AI 2.7 RP nasals 70 AI 2.8 RP fricatives 70 AI 2.9 RP affricates 71 AI 2.10 RP approximants 71 AI 2.11 Nasal versus oral sounds Audio Illustrations Chapter 3 AI 3.1 The spellings of RP /iː/. Further practice AI 3.2 The spellings of RP /ɪ/. Further practice AI 3.3 The spellings of RP /e/. Further practice AI 3.4 The spellings of RP /ɜː/. Further practice AI 3.5 The spellings of RP /ə/. Further practice

96 97 103 103 104

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:58 PM

xvi AI AI AI AI AI AI AI AI AI AI AI AI AI AI AI AI AI AI AI AI

3.6 3.7 3.8 3.9 3.10 3.11 3.12 3.13 3.14 3.15 3.16 3.17 3.18 3.19 3.20 3.21 3.22 3.23 3.24 3.25

Audio Tracks

The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings The spellings

of of of of of of of of of of of of of of of of of of of of

RP /ʌ/. Further practice RP /æ/. Further practice RP /ɑː/. Further practice RP /ɔː/. Further practice RP /ɒ/. Further practice RP /uː/. Further practice RP /ʊ/. Further practice RP /eɪ/. Further practice RP /aɪ/. Further practice RP /ɔɪ/. Further practice RP /əʊ/. Further practice RP /aʊ/. Further practice RP /ɪə/. Further practice RP /eə/. Further practice RP /ʊə/. Further practice RP /eɪə/. Further practice RP /aɪə/. Further practice RP /ɔɪə/. Further practice RP /əʊə/. Further practice RP /aʊə/. Further practice

Audio Illustrations Chapter 4 AI 4.1 The spellings of RP /p/. Further practice AI 4.2 The spellings of RP /b/. Further practice AI 4.3 The spellings of RP /t/. Further practice AI 4.4 The spellings of RP /d/. Further practice AI 4.5 The spellings of RP /k/. Further practice AI 4.6 The spellings of RP /g/. Further practice AI 4.7 The spellings of RP /f/. Further practice AI 4.8 The spellings of RP /v/. Further practice AI 4.9 The spellings of RP /θ/. Further practice AI 4.10 The spellings of RP /ð/. Further practice AI 4.11 The spellings of RP /s/. Further practice AI 4.12 The spellings of RP /z/. Further practice AI 4.13 The spellings of RP /ʃ/. Further practice AI 4.14 The spellings of RP /ʒ/. Further practice AI 4.15 The spellings of RP /h/. Further practice AI 4.16 The spellings of RP /ʧ/. Further practice AI 4.17 The spellings of RP /ʤ/. Further practice AI 4.18 The spellings of RP /m/. Further practice

109 110 111 115 116 119 120 129 130 130 134 134 138 139 139 143 143 144 144 144

169 170 176 177 182 182 187 187 191 191 197 198 201 202 204 209 209 214

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:58 PM

Audio Tracks

AI AI AI AI AI AI

4.19 4.20 4.21 4.22 4.23 4.24

The spellings The spellings The spellings The spellings The spellings The spellings

of of of of of of

RP /n/. Further practice RP /ŋ/. Further practice RP /r/. Further practice RP /l/. Further practice RP /j/. Further practice RP /w/. Further practice

Audio Illustrations Chapter 5 AI 5.1 Allophonic variation of vowels AI 5.2 Aspiration of voiceless plosives 245 AI 5.3 Assimilation 249 AI 5.4 Elision

xvii

214 215 222 222 227 227

236 239

Audio Illustrations Chapter 6 266 AI 6.1 Stress in polysyllables 266 AI 6.2 Variable stress AI 6.3 Suffixes which do not affect the pronunciation of the stem 269 AI 6.4 Stress shifting suffixes 270 AI 6.5 Stress carrying suffixes 274 AI 6.6 Prosodic stress 284 AI 6.7 Falling tones 289 AI 6.8 Attitudinal function of intonation 291 AI 6.9 Accentual function of intonation 292 AI 6.10 Grammatical function of intonation 299 AI 6.11 Discourse function of intonation 309 AI 6.12 Tag questions Audio Illustrations Chapter 7 326 AI 7.1 Pronunciation of the letter 334 AI 7.2 Pronunciation of the letter 336 AI 7.3 Silent letters AI 7.4 Pronunciation of the sequence 340 AI 7.5 Pronunciation of the -ed ending 341 AI 7.6 Some English homophones

268

339

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:58 PM

List of Abbreviations and Phonetic Symbols Abbreviations af bf C CV CVS dB EFL EPD F F0 F1 F2 Fr GA Hz IPA Lat L1 L2 LPD msecs n NLM NP NRP OBR OLD PAM pl post-alv PSp PresE

Meaning after/following before/followed by consonant Cardinal Vowel Cardinal Vowel Scale decibel(s) English as a Foreign Language (Cambridge) English Pronunciation Dictionary formant(s) fundamental frequency first formant second formant French General American Hertz International Phonetic Alphabet Latin First language Second language Longman Pronunciation Dictionary milliseconds noun Native Language Magnet normal pronunciation non-regional pronunciation openness, backness and lip rounding Oxford Learner’s Dictionary Perceptual Assimilation Model plural post-alveolar Peninsular Spanish Present-day English

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:47 PM

List of Abbreviations and Phonetic Symbols

RP SFS SID sg Sp SR SSBE SSE SSLE usu V v VT

xix

Received Pronunciation Speech Filing System Speech Internet Dictionary singular Spanish Speech Recognition Standard Southern British English Standard Scottish English Spanish-speaking learners of English usually vowel verb Vocal tract

Phonetic Symbols

Meaning

Examples

a

Cardinal Vowel no. 4 (open front unrounded)

First element of the RP diphthong /aɪ/ in idle [ˈaɪdl ̩]

ɐ

near-open central vowel

Scottish English pronunciation of the vowels in bud or putt

æ

raised open front vowel

in cat [kæt]

ɑ

Cardinal Vowel no. 5 (open back unrounded)

Fr pas ‘but’

ɑː

long open back unrounded vowel

in card [kɑːd]

ɒ

Cardinal Vowel no. 13 (open back rounded)

in pot [pɒt]

b

voiced bilabial plosive

in bib [bɪb]

β

voiced bilabial fricative

in PSp ave [ˈaβ̞e] ‘bird’

ɔ

Cardinal Vowel no. 6 (open-mid back rounded)

First element of the RP diphthong /ɔɪ/ in boy [bɔɪ]

ɔː

long open-mid back rounded vowel

in port [pɔːt]

d

voiced alveolar plosive

in dear [dɪə]

ʤ

voiced postalveolar affricate

in bridge [brɪʤ]

ð

voiced dental fricative

in this [ðɪs]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:47 PM

xx

List of Abbreviations and Phonetic Symbols

e

Cardinal Vowel no. 2 (close-mid front unrounded)

in dress [dres]

ə

central unrounded vowel: “schwa” post-nuclear and pre-nuclear in banana [bəˈnɑːnə]

ɛ

Cardinal Vowel no. 3 (open-mid front unrounded)

in Fr père ‘father’

ɜː

long open-mid central unrounded vowel

in herb [hɜːb]

f

voiceless labio-dental fricative

in fair [feə]

ɡ

voiced velar plosive

in hug [hʌɡ]

h

voiceless glottal fricative

in hip [hɪp]

ɦ

voiced glottal fricative

in ahead [əˈɦed]

i

Cardinal Vowel no. 1 (close front unrounded)

in PSp isla [ˈisla] ‘island’ and RP /i/ in noisy [ˈnɔɪzi]

iː

long close front unrounded vowel

in seed [siːd]

ɪ

lax close front unrounded vowel

in this [ðɪs]

ʝ

voiced palatal fricative

in PSp [ˈʝeso] ‘plaster’

k

voiceless velar plosive

in cap [kæp]

l

voiced alveolar lateral approximant

in eleven [ɪˈlevn̩]

ɫ

velarised voiced alveolar lateral/ dark l

in peel [piːɫ]

m

bilabial (realisation of a) nasal

in mother [ˈmʌðə]

ɱ

labio-dental (realisation of a) nasal in comfort [ˈkʌɱfət]

ɯ

Cardinal Vowel no. 16 (close back unrounded)

RP /uː/ with spread lips

n

alveolar nasal

in plan [plæn]

ŋ

velar (realisation of a) nasal

in sing [sɪŋ]

ɲ

palatal nasal

in PSp caña [ˈkaɲa] ‘cane’

o

Cardinal Vowel no. 7 (close-mid back rounded)

in PSp no [no] ‘not’

ø

Cardinal Vowel no. 10 (close-mid front rounded)

in Fr peu ‘little’

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:47 PM

List of Abbreviations and Phonetic Symbols

xxi

œ

Cardinal Vowel no. 11 (open-mid front rounded)

in Fr peur ‘fear’

Œ

Cardinal Vowel no. 12 (open front rounded vowel)

It has not been found to exist as a separate phoneme.

θ

voiceless dental fricative

in thin [θɪn]

p

voiceless bilabial plosive

in pet [pet]

r/ɹ

voiced post-alveolar approximant

in red [red] or [ɹed]

r

voiced alveolar trill or multiple vibrant

in PSp carro [ˈkaro] ‘cart’

ɾ

voiced alveolar tap

in PSp caro [ˈkaɾo] ‘expensive’

s

voiceless alveolar fricative

in miss [mɪs]

ʃ

voiceless postalveolar fricative

in ship [ ʃɪp]

t

voiceless alveolar plosive

in tin [tɪn]

ʧ

voiceless postalveolar affricate

in choose [ʧuːz]

u

Cardinal Vowel no. 8 (close back rounded)

in PSp útil [ˈutil] ‘useful’

uː

long close back rounded vowel

in food [fuːd]

ʊ

lax close back rounded vowel

in put [pʊt]

v

voiceless labio-dental fricative

in very [ˈveri]

ʌ

Cardinal Vowel no. 14 (open-mid back unrounded)

in mud [mʌd]

w

voiced labial-velar central approximant semivowel

in white [waɪt]

x

voiceless velar fricative

in PSp jarra [ˈxara] ‘jar’

y

Cardinal Vowel no. 9 (close front rounded)

in Fr du ‘from’

ʎ

voiced palatal lateral

in PSp llama [ˈʎama] ‘flame’

ɤ

Cardinal Vowel no. 15 (close-mid back unrounded)

Realisation of PresE /ʊ/ in some dialects

z

voiced alveolar fricative

in zoo [zuː]

ʒ

voiced postalveolar fricative

in measure [ˈmeʒər)]

ʔ

glottal stop

in button [bʌʔn̩]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:47 PM

xxii

List of Abbreviations and Phonetic Symbols

ʘ

bilabial click

!

alveolar click

ǁ

velar click

Diacritics

Meaning

Examples

ʰ [pʰ]

aspirated

in peel [pʰiːɫ]

⁼ [p⁼]

unaspirated

in spider [ˈsp⁼aɪdər]

voiced

in matter [ˈmæt̬ər]

˚ [ŋ̊] [d̥]

voiceless or devoiced

in did [d̥ɪd̥]

ʲ [kʲ]

palatalised consonant

and in keen [kʲiːnʲ]

̬ [ t̬]

ʷ [pʷ]

labialised consonant (lip rounding)

and in put [pʷʊtʷ]

̪ [d̪]

dental

in PSp monte [ˈmõn̪te] 'hill'

ˡ [dˡ]

lateral release

in middle [ˈmɪdˡl ̩]

ⁿ [tⁿ]

nasal release

in catnap [ˈkætⁿnæp]

nasal release

in topmost [ˈtopᵐməʊst]

˜ [ẽ]

nasalised

in ten [tẽn]

˺ [g˺]

non-audible release

in big [bɪɡ˺]

ˆ [pˆ]

unreleased

First

in top post [ˌtʰɒpˆ ˈpʰəʊst=]"

advanced (consonants)

in key [k̟iːː] or [k˖iːː]

advanced (vowels)

Quality of RP [uː]

retracted (consonants)

in pool [p̄ ːu ɫ] or [p-ːuɫ]

̙ [ɑ̙]

retracted (vowels)

[ɑ̙ː] retracted variant of RP [ɑː]

̈ [ɑ̎]

centralised

Quality of RP [ɑː]

˕ [ɪ ]̞

lowered or more open

in sit [sɪ ̞t]

˔ [ɪ ]̝

raised or closer

in bit [bɪ ̝t]

̯ [i ]̯

non-syllabic

in PSp aire [ˈai ̯re] 'air'

̩ [n̩]

syllabic

in eleven [ɪˈlevn̩]

ːː

extra-length

in bee [biːː]

ˑ

half-length

in beat [biˑt]

ː

normal length

in cord [kɔːd]

ˈ

primary stress

computer [kəmˈpjuːtər]

ˌ

secondary stress

understand [ˌʌndəˈstænd]

m

[pm]

˖ [k̟] ̘ [u̘ ] ˉ p̄

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:47 PM

List of Abbreviations and Phonetic Symbols

//

phonemic transcription

post /pəʊst/

[]

allophonic/phonetic transcription

post [pʰəˑʊst=]

*

ungrammatical/wrong pronunciation

earliest *[ˈɜːlɪst]

.

syllable boundary

PSp reyes [re.ιes] ‘kings’

|

tone unit boundary

||

pause

_#

word-finally

#_

word-initially

xxiii

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:47 PM

Purpose and Scope of the Book English Pronunciation for Speakers of Spanish. From Theory to Practice (EPSS) is aimed at meeting the needs of speakers of Spanish who want to learn or teach English phonetics and phonology at universities and teacher-training institutes, or otherwise wish to improve their English pronunciation and their skills in transcribing English phonetically. To be covered in one or preferably two semesters, the volume can be used in a course on English Phonetics and Phonology and/or on English Pronunciation. In addition, the book may be useful to anyone interested in gaining insight into the differences and similarities that exist between English and Spanish pronunciation to prepare the ground for more advanced and extensive reading in the field. EPSS is supported by a companion website called EPSS Multimedia Lab available at http://www.usc.es/multimlab/index.html. It contains the audio files (audio illustrations or audio exercises) accompanying the book, with entries for each numbered soundtrack in its corresponding unit and tab, sampling the voices of five native speakers of British English (three female and two male) and one female speaker of Peninsular Spanish, as well as animations, videos and additional material that can be used in combination with this manual or independently, including: (1) an Animated Sound Bank of English-Spanish (with phonemic transcriptions and original recordings), (2) Glossaries and Dictionaries of Phonetics and English Pronunciation, (3) Downloadable and Recorded Exercises (with their keys), and (4) Other Resources to Teach and Learn English Phonetics and Pronunciation. The term pronunciation in the title is a cover term for the contents of both the book and the website. EPSS provides a down-to-earth introduction to the basic principles, most significant concepts and terminology of English phonetics, adopting an essentially practical contrastive approach. Our intention is to show how practical phonetics can be effectively used both to learn English phonetics and pronunciation and to teach it to non-natives, in particular to Spanish-speaking learners of English (SSLE) by helping them to represent, perceive and reproduce the sounds of English as compared to those of Spanish. To this effect, the sounds of British English, more specifically the Received Pronunciation (RP) accent, are contrasted with those of Spanish, particularly Peninsular Spanish (PSp). Separate units are devoted to the discussion of vowels, vowel glides, consonants, and features of connected speech, including such phenomena as co-articulation, assimilation, elision, linking and prosody (stress, rhythm, intonation), as well as to the description of sound/ spelling relationships in English. In addition, the volume also offers SSLE

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

Purpose and scope of the book

xxv

guidance on how to pronounce and talk in conversations. To provide such guidance is of paramount importance to us, because after years of experience as teachers of English phonetics to Spanish university students we agree with Coe (2001: 91) that “European Spanish speakers, in particular, probably find English pronunciation harder than speakers of any other European language”. Accordingly, taking a communicative approach, the book highlights the phonetic and phonemic contrasts and specific cues that are most important to aiding comprehension in English. Likewise, the features of English pronunciation that are potentially problematic for speakers of Spanish are emphasised so as to prevent misconceptions and avoid, whenever possible, the presence of a foreign accent, which may result from a variety of factors, ranging from language universal and (L1 and L2) language-specific constraints (taking into account the learner’s native language or linguistic variety and its linguistic similarity to the target accent) to individual-dependent characteristics (age, instruction, phonetic and auditory abilities, as well as such affective factors as attitude, identity and concern for good pronunciation) (Kenworthy 1987; Morley 1991; Lecumberri 1999; Moyer 1999). As we are aware that these factors mean that SSLE are unlikely to start on the study of phonetics with a native-like pronunciation but are likely to carry the signature of the phonological structure and “articulatory setting” (position of articulators) of Spanish, specific sections of this book are devoted to serving as a guide towards “correct” pronunciation habits. Our purpose is to help SSLE sound as close as possible to native English or, at least, acquire an intelligible RP pronunciation (being able to produce the sound patterns of English). By achieving comprehensibility (so that the meaning of what is said is understood) and interpretability (so that the purpose of what is said is understood), they will satisfy their communicative needs in any situation (functional communicability) and increase their selfconfidence while also fostering their monitoring abilities and speech strategies (Burn 2003). Nevertheless, it should be noted that intelligibility may be affected by other factors besides pronunciation (Fayer and Krasinski 1987), such as grammatical correctness (Varonis and Gass 1981), the fluidity and rhythm of speech (Anderson-Hsieb and Kehler 1988), familiarity with the topic discussed (Gass and Varonis 1984) and background or ambient noise (Munro 1998). The book contains seven chapters. Each begins with a general introduction to the topic named in its title, followed by detailed analyses of the relevant issues, which are regularly interspersed with illustrative audio illustrations, examples and diagrams (e.g. waveforms, spectrograms, midsagittal sections of facial diagrams, Tables, Figures and other kinds of artwork). At the end of each chapter, there are also sections with recommendations for Further Reading and a battery of Exercises of different kinds. The proposed activities are devised

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

xxvi

Purpose and scope of the book

not only to help the reader gain practice in ear-training, oral production and phonetic transcription using the symbols and diacritics provided by the International Phonetic Alphabet, but also to encourage scientific thinking about phonetic issues. Although all the chapters can be read in any order, we recommend following the order given, since both the explanations and the exercises are cumulative in that later chapters are based on the contents explained in earlier ones so that previously presented technical terms are used without any further explanation. Chapter 1 gives an overview of the fields of phonetics and phonology, paying particular attention to the notions of phoneme and allophone and to the structure of the syllable and syllabic sounds. After this, the sounds of RP are presented as a pronunciation model for SSLE in comparison with those of PSp. In addition, the main differences between broad (or phonemic) and narrow (or allophonic) transcriptions are noted, and practical tips are given on how to represent speech sounds more accurately in transcriptions than would be possible using ordinary spelling, bearing in mind the complex and often unpredictable relationship between spelling and pronunciation. Chapter 2 summarises the principles of articulatory phonetics, describing the process that takes place between our lungs and our lips in the production of speech sounds. The organs of speech are described in relation to the function each plays in the articulation of speech sounds, noting that these organs are also involved in the realisation of other primary biological functions (e.g. breathing, licking and biting), as well as in the production of noises which are not speech sounds (e.g. coughing, sneezing and whistling). Descriptions are also provided of the speech organs and the air-stream mechanisms that are used to produce (pulmonic, glottalic and velaric) speech sounds, noting the action of the vocal cords that is responsible for the distinction between voiced and voiceless sounds. The chapter closes with a characterisation and classification of speech sounds in terms of their articulatory and acoustic features. Chapter 3 describes RP pure vowels and complex vowels (or vowel glides) in comparison with those of PSp, while Chapter 4 focuses on consonants, distinguishing four main groups: plosives, fricatives, affricates and approximants. In these two chapters each RP sound is described according to nine parameters: (1) IPA symbol, (2) Identification, (3) list of Allophones, (4) articulatory Description, (5) Environment and main allophonic realisations, offering a description of the main contexts of appearance of the sound in question as well as of its main realisations, (6) Spellings, (7) Regional and social variants, mentioning the most important alternative pronunciations of each sound in RP and other accents in order to show that pronunciation is not monolithic (see § 1.5), (8) Comparison with Spanish and advice, providing pronunciation tips that are essential for

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

Purpose and scope of the book

xxvii

intelligibility or otherwise may be relevant to the imitation of a native-like accent, and (9) Further practice, which includes recorded (and transcribed) examples (extracted from our Sound Bank). The exercises included in these two chapters are specially devised to facilitate improvement targetting the potential weaknesses that speakers of Spanish may have to improve in their production and perception of RP sounds (e.g. difficult or subtle phonemic contrasts, sounds that exist in English but not in Spanish, and so on). Chapter 5 explains common phenomena of connected speech in English: coarticulation, assimilation, elision, linking, juncture and gradation, the latter referring to the different realisations of sounds (i.e. weak vs. strong) that result from the metrical structure of speech and their position in the syllable. The chapter summarises the main variants (or allophones) of RP phonemes arising from the aforementioned phenomena with their corresponding notation conventions or diacritics, which show how sounds influence one another when put together in words, phrases, sentences and speech sequences. Roughly, what happens is that the faster we speak, the less carefully and clearly we distinguish the beginning of one word from the end of the previous one, and some features of final and initial sounds start merging together because of economy of articulatory effort (i.e. laziness). In close connection with the above mentioned phenomena, Chapter 6 concentrates on stress, rhythm and intonation. We shall see how words are stressed in isolation and in the stream of speech. Likewise, tonality, tonicity and tone are examined closely as the main constituents of intonation, and their functions are explained in detail. Lastly, Chapter 7 explores both the systematic and the unsystematic relationships between the written and spoken forms of English words. We believe it is important to know at least the main spelling patterns of each of the English phonemes because they can be used as predictors for pronunciation, although exceptions to general tendencies are also pointed out. The Answer Key at the end of the book provides sample answers for the written exercises included in each chapter, as well as the phonetic transcriptions of the texts that are proposed as further transcription practice. These are complemented, as already noted, with the (audio) activities (and their keys), as well as the resources that are available on the EPSS Multimedia Lab. The volume closes with a Reference section listing the sources that are either cited throughout the text or included in the Further Reading Section of each chapter. Given its practical orientation, the theory in the book has been kept as simple and accessible as possible. Our comparison of English and Spanish is based on prior literature but also on our own experience and observation over the years as university teachers of this subject. We accept responsibility for any

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

xxviii

Purpose and scope of the book

weaknesses and errors in this respect. To find out more about the issues raised in EPSS the reader is referred to such textbooks as Gimson’s introduction to the pronunciation of English (Cruttenden 2014), English phonetics and phonology: an introduction (Carr 2012), Understanding phonetics (Ashby 2011), Practical phonetics and phonology (Collins and Mees 2009), An introduction to phonetics and phonology (Clark et al. 2007), Speech sounds (Ashby 2006), English phonetics and phonology (Roach 2005), and A manual of English phonetics and phonology (Skandera and Burleigh 2005), to mention but a few. Other accounts of English phonetics and pronunciation that are specifically addressed to SSLE can be found in La pronunciación del inglés británico simplificada (Monroy Casas 2012), Teach yourself English pronunciation (Estebas Vilaplana 2009), English phonetics and phonology for Spanish speakers (Mott 2005), Fonética inglesa para españoles (Alcaraz and Moody 1993), La pronunciación del inglés RP para hablantes de español (Monroy Casas 1980), A course in English phonetics for Spanish speakers (Finch and Ortiz Lira 1982), Manual de pronunciación inglesa comparada con la española (Sánchez Benedicto 1980), The sounds of English and Spanish (Stockwell and Bowen 1965), and Una comparación entre los sistemas fónicos del ingles y del español (Lado 1965); while studies contrasting English with other languages are, for instance, The phonetics of English and Dutch (Collins and Mees 2003) and Comparing the phonetic features of English, French, German, and Spanish (Delattre 1965). Turning to texts on Spanish phonetics and phonology, we recommend Manual de fonética española. Articulaciones y sonidos del español (Martínez Celdrán and Fernández Planas 2007), Tratado de fonología y fonética españolas by Quilis (1993, 1985), Estudios de fonología española by Navarro Tomás (1966 [1946], 1991[1918]) and Fonología española (Alarcos LLorach 1961 [1983]). With respect to the phonetic transcription system used, it is very similar to that utilised by Peter Roach, James Hartman and Jane Setter in their different editions of Daniel Jones’s (Cambridge) English Pronouncing Dictionary (EPD) (2011; 15th, 16th and 17th edns. 1997, 2003, 2006) and by J. C. Wells in his Longman Pronunciation Dictionary (LPD) (2008; 1st and 2nd edns. 1990, 2000). If you are interested in transcription manuals in particular, we recommend Tench (2011), Lillo (2009), Monroy Casas (2001) and García Lecumberri and Maidment (2000), but if what you want is to practise English pronunciation, some classic books are Baker (2007), Arnold and Gimson (1976) and Gimson (1980). EPSS provides transcriptions of English words that are frequent and/or somehow troublesome for foreign learners, as well as passages that illustrate the common processes of connected speech, with the intention of providing a natural and accurate representation of how native speakers sound. Phonetic transcription is necessary to raise awareness of the target that should be aimed

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

Purpose and scope of the book

xxix

at, as well as of the possible pronunciations to be expected from native speakers. But this is a skill that can be best mastered by regular practice. So, whether working alone or in a group (where you can learn from others), SSLE are encouraged to engage in the task itself as often as possible, first transcribing sets of words and then checking their progress with the help of the answers provided in the Appendix and the website, before moving on to more advanced transcriptions. It should be borne in mind, however, that the transcription keys provided are only model answers. This means that they are always acceptable, but not necessarily the only possible ones as there may be other alternatives since pronunciation is subject to more variation than any other aspect of language. Finally, on the issue of speech analysis, we have used Speech Filing System (SFS 4.7 and SFS/WASP Version 1.54) for spectrographic and waveform analysis (Huckvale 2008), freely available at http://www.phon.ucl.ac.uk/resource/sfs/ wasp.htm, while PRAAT 5.0.20 may be used for intonation curve analysis (Boersma and Weenink 2008), which can also be freely downloaded at http:// www.fon.hum.uva.nl/praat/. Learners are advised to record their own pronunciation to use these audio files for reference and checking. Additional information and resources may be found in the corresponding sections of the EPSS Multimedia Lab.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 1:59 PM

Chapter 1

1 Phonetics and Phonology 1.1 Introduction If we simplify the distinctions within the whole of linguistic science we can say that basically there are two kinds of branches: external and internal ones. External branches of linguistics deal with disciplines that can influence or be influenced by language (e.g. sociolinguistics, ethnolinguistics, psycholinguistics, neurolinguistics, etc.). In contrast, internal branches are concerned with the study of different aspects of language itself, such as morphology (the analysis of morphemes), lexicology (the study of words), syntax (the study of how words are used to create phrases, clauses and sentences), semantics (the exploration of meaning), or pragmatics and discourse (the analysis of individual utterances and of organised sets of utterances or text units in context). Although in terms of methods they are certainly very different from the rest, phonetics and phonology can be regarded as two further interrelated internal fields of linguistics in that, broadly, both explore the phonic or sound component of language, an indispensable foundation of linguistic enquiry.1 We do believe that the boundary between phonetics and phonology need not be sharply drawn, nor should it be constructed on assumptions about the primacy of one over the other. For, although we analyse speech by breaking it down into its several aspects, the reality is one of integration because speech sounds cannot be thoroughly studied exclusively in isolation without looking at their linguistic function and context, in much the same way that their function cannot be properly analysed without considering their articulatory and/or acoustic features (Lass 1984: 1). However, we shall distinguish phonetics from phonology for methodological reasons and because it seems true that phonetics can be studied without really exploring phonology, while phonology is closely dependent on phonetics for the data on which it relies to pursue its arguments (Gussenhoven and Jacobs 2005). Ultimately, what this book shows is that one thing is to speak without deliberately pondering on what we are doing, and another is to study this process systematically for the purpose of its scientific analysis, as it happens in phonetics and phonology, two fields of linguistics that are being constantly updated and extended. 1 Other dimensions of phonetic and phonological analysis are: (1) the combinatory possibilities of the sounds (the phonotactics or syllable structure, see § 1.3.3); (2) the prosody of the language (pitch, loudness, length, accent, rhythm, intonation, see Chapter 6); and (3) the relationship between sounds and letters (graphology or graphemics, see Chapter 7).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

2

Phonetics and Phonology

1.2 Phonetics Often depending more on data, scaled measurements, instruments and technologies than other areas of linguistics, phonetics is an empirical science (based on the observation of facts) which studies the concrete characteristics of human sound-making, especially speech sounds2 but also involuntary noises (e.g. hiccups and coughing) as well as other aspects of voice. For the phonetician, sounds are phenomena in the physical world and phonetics provides information on their physical properties, allowing us to devise methods of sound description, classification and transcription (Crystal 2008: 363–364). Phonetic categories are thus generally defined using terms which have their origins in external disciplines such as anatomy, physiology and acoustics, so that consonant sounds, for example, are usually described by making reference to anatomical place of articulation (dental, palatal, etc.) and their physical makeup (the frequency and amplitude of consonantal sound waves). The discipline is often referred to as general phonetics because it is based on the assumption that the methods of analysis are equally applicable to the sounds of any language in the world, which reflects the phoneticians’ attempts to discover the universal principles underlying the nature and use of speech sounds. Experimental phonetics is another term used to emphasise the “pure” scientific endeavour of general phonetics. General phonetics, interpreted as involving general studies of speech sounds, is usually distinguished from functional phonetics, which investigates the phonetic properties of specific languages or how the sounds are used within the pronunciation system of a language, an approach which is usually carried out under the heading of phonology, which will be further described in section 1.3. Three branches of phonetics are generally recognised depending on which phase of the speech chain (the process in which humans produce and hear speech) is being described: articulatory, acoustic and auditory. Our ability to communicate is apparently an easy and unremarkable action, but beneath the surface there are five complex processes at work labelled A to E in Figure 1 below (adapted from Denes and Pinson 1993: 5). Stages A and E involve the formulation and interpretation of the message in the brains of speaker and listener respectively, forming the link between phonetics and psycholinguistics. We are constantly monitoring our own speech by listening to our performance, a process termed audio-feedback, which is represented in the diagram as the 2 Speech sounds are a small subset of all the noises which humans can produce with their vocal apparatus in order to form words.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonetics

3

Figure 1: The speech chain

feedback link. Stage B symbolises the bodily function that is responsible for the production of speech sounds (articulatory phonetics). In Stage C the physical nature of speech sounds can be examined, as well as how they are transmitted through the air as sound waves (acoustic phonetics). Stage D corresponds to the way in which the addressee’s ear perceives the speech signal (auditory phonetics). We can thus conclude that phonetics analyses the anatomy and physiology of speech sounds, integrating the three aspects mentioned above, i.e. articulatory, acoustic and auditory/perceptual. The study of any of these aspects, involving appropriate instrumental analysis techniques (e.g. airflow measurement, speech synthesis, sound wave analysis, sampling, averaging, etc.) is covered by the umbrella term instrumental phonetics. Phonetics plays an important role in the teaching of foreign languages and is also useful in the acquisition of good diction, in speech therapy for people with speech and hearing impediments, as well as in sound transmission and forensic linguistics. In what follows, separate sections will be devoted to describing articulatory, acoustic and auditory phonetics. The emphasis of this book is mainly on articulatory phonetics, but whenever necessary, acoustic and/or auditory considerations are also incorporated.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

4

Phonetics and Phonology

1.2.1 Articulatory Phonetics Articulatory phonetics explores the nature and limits of the human ability to use the speech organs to articulate speech, despite organic differences, transforming aerodynamic energy into acoustic energy or sound waves, which are then perceived by the human auditory system as speech sounds (Laver 1994). Speakers differ organically from each other in anatomical factors such as the dimensions, mass and geometry of their vocal organs (e.g. the shape and size of the articulators and the speech cavities), and yet they may be judged to be producing linguistically and paralingistically identical utterances that are perceived as speech sounds. As a producer of speech sounds, you may be already aware, even if intuitively, of some aspects of articulatory phonetics. Now you need to deepen your awareness of the movements, gestures and feelings occurring in your vocal tract (VT) when producing speech sounds, and learn some specialised vocabulary to be able to verbalise such knowledge. For instance, we shall see that from an articulatory point of view the RP sound /p/ is a voiceless bilabial plosive because its production involves no vibration of the vocal cords, but a close compression of the lips and generally a palpable puff of air upon release from the mouth (see § 4.2.1.1). Articulatory phonetics is the branch of phonetics most widely taught, underpinning both acoustic and auditory phonetics, in such wide-ranging disciplines as languages and linguistics, speech and language therapy, medical science, voice and singing studies, and drama, among others (Ashby 2011: 9). The principles and details of articulatory phonetics are further detailed in Chapters 2 to 5 when examining the production of speech, English vowels and consonants in comparison with those of Spanish, and connected speech phenomena, respectively.

1.2.2 Acoustic Phonetics Acoustic phonetics studies the physical properties of speech sounds (e.g. frequency, amplitude, rate, etc.) as transmitted between mouth and ear (Crystal 2008: 7–8). Two instances of the same spoken sequence made by the same speaker are in fact most unlikely to be acoustically identical despite their phonetic likeness. Likewise, identical utterances from two different speakers are bound to be even more acoustically different. Acoustic measurements serve to support these and other articulatory or auditory judgements, the most widely used being those provided by waveforms and spectrograms. The waveform of a sound represents a moving airstream which is modified by the articulators as it travels through the VT. Displaced particles of air move

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonetics

5

away from rest position and back to it, displacing other particles and causing them to repeat the same vibratory pendulum-like movement. This pattern of movement has the shape of a wave, and its acoustic representation is called a sine or sinusoidal wave on the grounds of certain mathematical properties that it has. Figure 2 below plots examples of sine waves, where the x axis indicates time in seconds or milliseconds (.001 second or 1 millisecond), and the y axis relative intensity in arbitrary units. Waveforms may be periodic (as in Figure 2 (a) below) or aperiodic (as in Figure 2(b)). Aperiodic waveforms have no identifiable periodicity, and are characteristic of what we call noise (e.g. that produced by thunder, clapping hands and splashing water). Periodic waveforms, on the other hand, are more characteristic of speech, singing and birdsong, and can be described in terms of amplitude, intensity and frequency. The amplitude of the wave is the maximum distance that a particle moves in each direction from its starting point, that is the distance from the base line to the highest point of the curve. In general the greater the amplitude, the louder the sound is perceived to be. Amplitude correlates with intensity, which is measured in decibels (dB). In Figure 2 (a) below we can see that the two sine waves have identical amplitude and intensity (vertical axis). The frequency of a sound wave is the number of cycles (movement from rest position to position 2, back to position 3 and back to rest again) completed within a given limit of time (usually one second). Frequency is measured in Hertz (Hz): one Hz equals one cycle per second. A cycle can be measured from any point on a wave to the next point. The fundamental frequency (F0) of a sinewave is the number of times per second the vocal folds vibrate. At the same time as producing the fundamental frequency, a range of other higher frequencies called overtones or harmonics are also produced, which arise from the complex wave effects that occur when the vocal folds vibrate, a movement that can be viewed on stroboscopic images of vocal fold vibration and is often referred to as the mucosal wave (see § 2.2.2). Harmonics are multiples of the fundamental frequency and contribute to the resonant quality or timbre of the sound. The arrow in Figure 2 (a) points to where the oscillation repeats itself at 0.004 seconds. This indicates that one cycle of this wave lasts 0.004 seconds, so its F0 equals 1/0.004 seconds or 250 Hz. Then the second harmonic will be at approximately 500 Hz, the third at 750 and so on. These peaks of intensity are called formants (F) or VT resonances, and they are numbered upwards from the lowest in frequency. We shall see that the first two formants, F1 (first formant) and F2 (second formant) are related to vowel articulation (section 2.4.1). We perceive frequency as pitch. Pitch variation is produced primarily by stretching the length of the vocal folds which results in

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

6

Phonetics and Phonology

(a) Periodic sinewaves

(b) Aperiodic sinewaves

Figure 2: Periodic and aperiodic sinewaves

different intonation patterns and tonal distinctions on vowels. The higher the frequency, the higher the pitch. The range of pitch found in speech is about 60–500 Hz (men between 100 and 150 Hz, women between 200 and 325 Hz, and children around 265 Hz), although we do not usually use the entire range in speaking. In Figure 3 below, sinewaves (a) and (b) have the same frequency (200 Hz) but different amplitudes, low (a) and higher (b); while sinewaves (c) and (d) share

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonetics

7

approximately the same amplitude, but differ in frequency: (c) has a low frequency and (d) a much higher one.

Figure 3: Sinewaves with different amplitudes and frequencies

In addition, periodic waveforms can be simple sinewaves or complex ones, which consist of more than one sinewave and can be broken down into these component waves. In speech all periodic waveforms are complex, an illustration of which is presented in Figure 4 below, consisting of three sinewaves with different amplitudes and frequencies (viz. 100Hz with the greatest amplitude, 200Hz with the smallest amplitude, and 300Hz with a medium one) (Ladefoged 1996).

Figure 4: Constructing a complex wave

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

8

Phonetics and Phonology

Spectrograms, on the other hand (also known as spectral waterfalls, sonograms, voiceprints, or voicegrams), are time-varying spectral representations (i.e. forming an image) that show how the spectral density of a sound varies with time. The instrument that generates a spectrogram is called a spectrograph. Spectrograms have a three-dimensional display. The horizontal axis represents time, the vertical axis frequency, and a third dimension indicates the amplitude of a particular frequency at a particular time, which is represented by the intensity, or concentration of energy at particular frequency bands (the formants) that stand out in darker colour in the image. Figure 5 below illustrates the speech waveform and wideband spectrogram for one realisation of the utterance “passed London”.

Figure 5: Speech waveform and spectrogram of “passed London”

Speech spectrograms are called wideband or narrowband depending on what window length is used, as shown in Figure 6 below containing the spectrographic representation of /m/ in “machali” (adapted from Jawale 2010). The shorter the window, the larger its bandwidth. So if the window is shorter, the spectrogram is called wideband, and if the window is larger, narrowband. Wideband spectrograms are widely used in phonetics because they reflect a crucial feature of speech sounds, namely their formant structure (i.e. amplitude peak in the frequency spectrum of a sound); whereas narrowband spectrograms are less often used because they reveal harmonic structure (pitch), a kind of information which is of comparatively less importance to learners. Acoustic measurements like those provided in Figure 6 below can be obtained quite easily using a computer, a microphone and freely downloadable

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonetics

9

Figure 6: Narrowband and wideband spectrograms

software such as WASP or Praat. These programs allow us to process and analyse the acoustic recordings of speech. Increasingly today this kind of acoustic measurements and interactive displays are used in different disciplines such as forensics, psychology, speech therapy, pronunciation training, and language teaching, where interactive displays based on speech waveform analysis are increasingly used to assist specially foreign language learners in fine-tuning their pronunciations (Ashby 2011: 10), as in the case of Cabrera Abreu and Vizcaíno Ortega (2009). Chapter 2 offers acoustic representations of vowels and consonants (waveforms and spectrograms) and summarises the acoustic features of these speech sounds.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

10

Phonetics and Phonology

1.2.3 Auditory Phonetics Phonetics is just as much about how we hear or perceive what is said as about how we say what we say. Auditory phonetics investigates the perceptual response to speech sounds, as mediated by ear, auditory nerve and brain (Crystal 2008: 44). Although the mental processes linked to the perception of speech are still largely unknown, we do know that the human ear can only hear sounds having certain characteristics. So the main interest of auditory phonetics lies in determining the processes whereby speakers discriminate speech sounds based on the perception of: (1) their temporal characteristics (i.e. perceived length); (2) their prosodic attributes (pitch, tone and loudness); and (3) the way these facets are interrelated with rhythm and stress in conformity with the metrical structure of speech. Auditory phonetics underpins much of ear-training as well as practical phonetic training and accent coaching, which prove very useful for improving the pronunciation of a foreign language. So you would do well to try and identify (and reproduce) all the sounds of RP in comparison with those of PSp, as well as with the other sounds (and corresponding symbols) included in the IPA chart. At a more theoretical level, another dimension of auditory phonetics is the study of speech perception involving such aspects as the hearing mechanism and the effects of sounds on the brain which are explored by audiologists, psychologists, neurologists, and other specialists (Ashby 2011: 10). 1.2.4 The interface of the auditory, acoustic and articulatory levels The categories recognised in auditory phonetics have correlates in acoustic and articulatory phonetics. Thus, as shown in Table 1 below (adapted from Alcaraz and Moody 1993: 13), we can say that pitch is articulatorily related to the position of the vocal folds and to the different articulations of speech sounds, on the one hand, and acoustically to the different formant structure of the spectrographic representation of such sounds, on the other. Table 1: The interface of the auditory, acoustic and articulatory levels Articulatory Phonetics

Acoustic phonetics

Auditory phonetics

Articulators

Formants

Pitch

Vocal folds

Frequency (n° cycles; wave; Hertz)

Tone

Effort/Intensity

Amplitude (intensity)

Loudness

Quantity/Duration

Rate / Tempo (pace delivery)

Length

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

11

Likewise, tone is articulatorily related with the action of the vocal cords, and acoustically, with the frequency of vibration of these organs; while loudness is in principle associated with articulatory effort and, acoustically, with amplitude. Lastly, length or duration relates to articulatory quantity and to rate or tempo in acoustic phonetics. To summarise, remember that pitch, tone, loudness and length are psychological, perceptual characteristics, whereas formant structure, frequency, amplitude and rate are physical properties of speech sounds.

1.3 Phonology Phonology is a branch of linguistics that studies the systems and structures of speech (Crystal 2008: 365–366) and intends to show how sounds function in a systemic way in a given language (Cruttenden 2014: 3). The term system indicates that we operate with the finite options that are available in a given language. The significance of any particular selection within a system lies in the contrast between what is selected and what could have been selected. Accordingly, in a phonological system, for example, the choices are limited and make sense only by reference to the system itself, a point which has long been recognised in discussion of phonological distinctiveness (Ball and Quayle 2009). In contrast, the term structure suggests that choices are made within a “structured” scheme or framework, but sometimes it refers to the linear organisation of language. In this second sense, structure can be contrasted with system, reflecting the two dimensions of linguistic organisation that are often referred to as paradigmatic and syntagmatic, respectively. Paradigmatic relations are those that exist among the options in a system, for example between a phoneme and the other phonemes to which it is opposed. Syntagmatic relations, on the other hand, are linear or sequential, operative for example in the coarticulation or assimilation of adjacent sounds or in the organisation of alliteration or rhyme across longer stretches of language. Sound systems comprise the meaningful relations and distinctions that exist among speech sounds across and within languages. Speech sounds are organised into a system of contrasts, which are analysed in terms of phonemes, distinctive features, combinations of sounds (sound structures such as syllables, words, etc.) or other phonological units according to the theory adopted. As a result, phonology is often said to be marked by abstraction and generality. It investigates what properties of speech sounds have a functional, communicative value. The aim of phonologists is to demonstrate the existence of distinctive

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

12

Phonetics and Phonology

sounds and patterns in a language, to investigate their function, behaviour and organisation, and to spell out the principles and rules underlying phonetic relationships, making as general statements as possible about the nature of sound systems. If phonetics provides descriptions of sounds and ways of classifying them, phonology is a kind of functional phonetics which employs these data to classify the speech sounds of a language into a system of contrasts. For example, the English words, pan, tan, can, ban, Dan, fan, van and ran, illustrate a meaningbearing or contrastive opposition that is triggered by their initial consonants /p t k b d f v r/. These consonants create a phonological system because they stand in potential distinctive opposition to each other. Phonologists are interested in determining such oppositions and use such terms as distinctive, functional, contrastive or information-bearing to refer to them. Two branches of the subject are usually recognised: segmental and suprasegmental phonology. Segmental phonology analyses speech into discrete segments or sounds (e.g. (semi)vowels, (semi)consonants), while suprasegmental or non-segmental phonology analyses those features which extend over more than one segment, such as intonation patterns. A further distinction is made between diachronic and synchronic phonology: the former explores patterns of sound change in the history of language, while the latter investigates sound patterns regardless of the processes of historical change. In the following sections we shall address key issues in phonological theory: (1) the difference between phoneme and allophone (Section 1.3.1); (2) some basic principles of phonological analysis (Section 1.3.2); and (3) the basics of the syllable, its structure in English and Spanish and syllabic consonants (Section 1.3.3).

1.3.1 Phone, phoneme and allophone A phone (enclosed between square brackets, as in [p]) is a single phonetic segment, viewed in terms of its phonetic character without regard to its possible phonological status. The term “phone” is related to “phoneme” in the same way “morph” is related to “morpheme”. A phoneme (enclosed between slant brackets, as in /p/) is an abstract segment in the phonological system of a particular language or speech variety that is generally described as the smallest linguistic unit which can make a difference in meaning (Jones 1967/1976). The notion originated from the need to establish patterns of organisation out of the indefinitely large range of sounds that may be used in languages. Phonemes stand in constrastive distribution, that is, they can occur in the same phonetic context and if they do,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

13

they produce a change of meaning. For example /d/ and /t/ are phonemes in English because both can occur word-finally and substituting one by the other (substitution test) triggers a change of meaning as in e.g. bed – bet /bed/ – /bet/. Such pairs of words (usually with different meanings) that differ only in one phoneme are called minimal pairs. Closely related with this, another feature of phonemes is that they can be contrasted, that is, they can be characterised and opposed to or distinguished from the rest, in terms of the values they have as regards a limited set of phonological features (e.g. voice, place of articulation, manner of articulation, lip-rounding, back, central, etc.): different phonemes will have different phonological features. A third feature of phonemes is that they are limited or fixed in number. We can only speak of the phonemes of some particular speech variety or a particular accent of a given language, because the number of phonemes varies from one language to another. Thus, in section 1.5 (see Table 4) it will be shown that RP has forty-nine phonemes, and PSp forty-two.

AI 1.1 Identification of phonemes In addition, the notion of phoneme allows linguists to group together sets of phonetically similar phones as variants, or “members”, of the same underlying unit. Phones are said to be realisations of one phoneme, and the variants are referred to as allophones of the phonemes. Allophones (represented between square brackets, as in [pʰ]) are phonetically distinct realisations of a single phoneme that are contextually determined. There are two kinds of allophones: (1) intrinsic, when they occur spontaneously as a result of the phonetic context (such as the nasalisations of vowels when followed by a nasal consonant); or (2) extrinsic, which are produced systematically (without a physiological reason) by speakers of a certain language, as in the case of aspirated voiceless plosives at the beginning of stressed syllables in English (Tatham and Morton 2011). Whether intrinsic or extrinsic, the allophones of a phoneme display three main characteristics: (1) they exhibit phonetic similarity, that is, they show a similar phonetic make-up that allows us to relate them as variations of a particular phoneme; (2) they do not entail a change of meaning; and

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

14

Phonetics and Phonology

(3) they may stand in either complementary distribution or in free variation. Allophones are in complementary distribution when they occur in different, or mutually exclusive, phonetic environments (e.g. aspirated and unaspirated plosives in English); and they stand in free variation when they occur in the same phonetic context, usually as a result of regional, register or social pronunciation variants (see § 1.3.2). To summarise, the phoneme can be described as an indivisible and minimally abstract structureless unit that has identifiable phonetic characteristics and may be realised in speech by phonetically different phones (its allophones) in the same or in different environments (Trask 2012). It is in this sense that phonemes may be viewed as a set of sounds: each member of this set is an allophone of the phoneme. The allophones of a phoneme, on the other hand, are united within it by their shared phonetic similarity and they may stand either in complementary distribution or in free variation. The most important thing for communication is to be able to recognise and use not only the full set of phonemes of the language being used, but also its allophones, particularly those that stand in complementary distribution as they can cause intelligibility problems or otherwise can result in a foreign accent. For instance, an untrained English speaker can tell by simply listening that the two instances of [l] in little are not the same (clear [l] vs. dark [ɫ]), and yet s/he is aware that they correspond to one l-sound in English. Likewise, a Spanish speaker may not be aware of the fact that the two occurrences of [d] in the word dedo ‘finger’ are phonetically different ([dental plosive [d̪] vs. dental fricative [ð]), but s/he will certainly tell that the two phones correspond to one dsound in Spanish. This can only be adequately explained by considering how speech sounds are produced (phonetics) in combination with how they function in a given language (phonology), which reinforces the idea presented in the Introduction, that phonetics and phonology are integrated.

1.3.2 Phonological analysis A basic principle of phonological analysis consists in distinguishing phonemes from allophones (Crystal 2008: 361–362). By way of illustration, we can say that in RP /t/ and /p/ are two consonantal phonemes because they stand in parallel or contrastive distribution, as illustrated by a minimal pair such as tin – pin: if /t/ is replaced by /p/ (or vice versa) (substitution test) at the beginning of

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

15

the segment -in, then we obtain two different English words, which have two different meanings.

AI 1.2 Contrastive distribution of phonemes in RP In addition, in English /t/ has different allophones or realisational variants depending on the phonetic context in which this phoneme occurs and which therefore are said to be in complementary distribution, such as [tʰ] (‘aspirated’), [t⁼] (‘unaspirated’), [tⁿ] (‘nasalised’), [t̬] (‘voiced’), etc. We could then work out a phonological realisation rule that indicates the actual pronunciation of this phoneme in a number of contexts:

Figure 7: Realisation rule for some allophones of /t/

In Figure 7, the arrow head following the curly bracket means ‘is realised as’ or ‘is pronounced as’, while the slant line (/) separates the operation of the rule (to the left of it) from the environment (to the right), and the underscore (or underdash) (_) represents the position of the sound itself in relation to its environment. In addition, the “number sign”, [#], indicates a word boundary; therefore (#_) means ‘immediately after a word-boundary’ and (_#) ‘immediately before one’. Finally, [V_V] means ‘between two vowels’, and we can say “elsewhere” to cover all the other phonetic environments that fall outside the environments specified in a phonological rule. In section 1.4 more indications will be provided to help you transcribe English phonemes and their allophones. For any particular system biuniqueness is a requirement. This implies that phones, phonemes and allophones can be unambiguously assigned to each other by means of general rules or strategies of pronunciation rather than by idiosyncratic adjustments to individual phonemes. Phonological relationships, however, are not always biunique. One phenomenon contributing to this lack of biuniqueness is neutralisation. A neutralisation occurs when two (or more) closely related phonemes that are in contrast with each other in most positions are non-contrastive in certain other positions. When this happens, the “opposition” between the two phonemes is said to be “neutralised”. In English, for example, the voicing contrast is neutralised after initial [s-]. By way of illustration, [t, d] contrast in most environments: initially (tip – dip [tɪp, dɪp]), finally

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

16

Phonetics and Phonology

(cat – cad [kæt, kæd]), after [l] (colt – cold [kɒlt], [kəʊld]), and after nasals (shunt – shunned [ʃʌnt, ʃʌnd]). But after initial [s-], the voicing opposition is neutralised because only a voiceless (or devoiced) consonant may occur in this context, and as a result the contrast between [t] and [d] in that phonetic context does not exist, nor, similarly, is there a contrast between [p] and [b] or [k] and [g] in this environment (sting [stɪŋ], – *sdin]). A further neutralisation of the phonemic contrast between [s] and [z] can be observed in the formation of English plurals. After voiceless sounds, only the voiceless fricative [s] is possible (e.g. [-ps] ropes, [-ts] bits, [-ks] docks, [-fs] waifs, [-θs]) breaths); whereas after voiced sounds, only the voiced counterpart [z] occurs (e.g. [-bz] robes, [-dz] bids, [-gz] dogs, [-vz] waves, [-ðz] lathes). The [s] – [z] contrast is also neutralised in verb-endings (e.g. laughs, waits, lives) and in possessives (e.g. John’s, Jack’s). But in this case the environment is restricted to obstruents,3 since [s, z] may still contrast after a sonorant 4 (e.g. pence – pens [pens, penz], since – sins [sins, sinz], else – sells [els, selz]). The effect of a neutralisation, then, is the narrowing of a set of contrasts in a particular environment. There exist three main types of neutralisation. The first takes place when a language has a certain contrast but only one of the relevant phonemes in contexts of neutralisation. This is the case of English voiced [b, d, g] and voiceless plosives [p, t, k]. They occur in word-initial and word-medial positions, but only voiceless or devoiced plosives occur word-finally. Therefore, it can be concluded that the contrast of voicing is inoperative or neutralised word-finally (e.g. tack [tæk] – tag [tæg̊]). The second type of neutralisation may be represented by some kind of variation or alternation among the otherwise contrasting phonemes. For example, in English a nasal consonant tends to be homorganic (i.e. produced in the same place of articulation) with a following stop within a word, and so we can say that when this happens the contrast with other non-homorganic nasals is neutralised in this context (nasal + plosive). Take, for example, the case of the homorganic series bilabial nasal and plosive [mb, mp] (combat, implausible). In this context the contrast between [m] and [n] is neutralised because non-homorganic series [nb, np] are not possible (*conbat, *inplausible). Alternatively, neutralisation may be represented by free variation, or “either or relations”. Free variation occurs when two (or more) sounds or phones appear in the same environment (parallel distribution) without a change in 3 Obstruent is cover-term for sounds produced with a constriction which impedes the flow of air through nose or mouth, as in plosives, fricatives and affricates (Crystal 2008: 338). 4 Sonorant sounds are defined articulatorily, as those produced with a relatively free airflow and vocal fold vibration (voiced), as in vowels, liquids, nasals and laterals (Crystal 2008: 442).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

17

meaning and without being considered incorrect by native speakers, as in alternative pronunciations of the same words in regional, register or social variants (e.g. either [eɪðə] vs. [iːðə]) or in individual speakers’ variants (“prestige”, personal style, pathologies, etc.) (e.g. as when speakers articulate the word pit with a released or an unreleased plosive) (see Sections 4.2.1.1 and 5.2.6 for a discussion of the types of plosive release and released sounds, respectively). The phonemes and realisations of phonemes (allophones) that stand in parallel distribution are called free variants as opposed to contrastive variants and contextual variants which are found in cases of contrastive (phonemes) and complementary distribution (allophones), respectively (see § 1.3.1). In terms of phonological analysis, there are three ways of treating neutralisations: (1) to abandon biuniqueness (no indeterminacy when invoking grammatical or semantic relationships); (2) to insist that sounds representing neutralisations must be treated as allophones of a phoneme (e.g. word-final voiceless plosives are taken at face value): voiced plosives do not occur word-finally; or (3) to recognise the notion of archiphoneme, which consists of the shared features of two (or more) closely-related phonemes, but excludes the feature which distinguishes them (Trubetzkoy 1969: 79). Thus the archiphoneme of [p b] consists of the features [bilabial and plosive] (but excludes voicing, which separates them); the archiphoneme of [s z] is [alveolar, fricative]; the archiphoneme of [m n ŋ] is [nasal]; and so on. For transcription purposes, the symbols [P T K] can be used to represent the archiphonemes of [p b], [t d] and [k g], respectively. The concept of neutralisation must be distinguished from a superficially similar phenomenon known as defective distribution. Defective distribution means that a sound does not occur in all possible environments and as a result its distribution is said to be “incomplete”. Three examples of sounds showing defective distributions are RP [ŋ], [h] and [f]. [ŋ] may follow a short vowel, but not a long vowel, and it can occur finally, but not initially. On the other hand, [h] can occur initially in a syllable, but not finally, and it can precede a glide consonant ([hj-, hw-]), but not a liquid5 (*[hl-] *[hr-]); while [f] can occur initially ( fine), medially (wafer) and finally (knife), but not before another obstruent word-initially (e.g. *[fp-], *[ft-]).

5 The term liquid is used collectively to refer all the apico-alveolar sounds (made by placing the tip of the tongue in the alveolar ridge) of the types [l] and [r] (Crystal 2008: 286).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

18

Phonetics and Phonology

The following series of words illustrate the realisation of the phonemic contrasts between English /t/ and /d/ and Spanish /m/ and /n/ in various word positions, as well as the neutralisation processes that occur both in English and Spanish concerning /t/ – /d/, /p/ – /b/, /k/ – /g/ and /m/ – /n/. English: /t/ – /d/ as distinct phonemes: tip – dip; cat – cad; colt – cold; shunt – shunned. /t/ – /d/ as neutralised phonemes: strain – its drain. /p/ – /b/ as neutralised phonemes: speak – its beak. /k/ – /g/ as neutralised phonemes: score – it’s gore. Spanish: /m/ – /n/ as distinct phonemes: mapa ‘map’ – napa ‘synthetic leather’, mamá ‘mom’ – maná ‘manna’. /m/ – /n/ as neutralised phonemes: hoNbre ‘man’.

1.3.3 Phonological structure Most native speakers of a language can usually give a quick answer if asked to count the number of syllables in a given word. However, although this division seems quite simple and straightforward, it is very frequent to find disagreement among speakers on the syllabic division of certain “problematic” words, and problems also arise when trying to provide a definition of this linguistic concept that is unanimously agreed upon. Section 1.3.3.1 offers a brief summary of the main definitions of the syllable. After this, Section 1.3.3.2 gives an overview of the phonotactic possibilities of English and Spanish, that is, the vowel-consonant combinations which are possible in syllable-initial and syllable-final positions in these two languages, and it also provides advice to SSLE on the acquisition of English consonant clusters. Closing our description of phonological structure, Section 1.3.3.3 focuses on a specific syllable type, that having a syllabic consonant as a nucleus.

1.3.3.1 The syllable Words can differ from each other not only in terms of which sounds they consist of, but also in terms of the particular sequence of consonants and vowels that gives them their structural shape. Languages have different restrictions on the combinatory possibilities and distribution of speech sounds within words,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

19

and this is the major concern of phonotactics: a discipline of linguistics that explores the combinatorial constraints and possibilities within syllable structure, as already noted in the introduction. The syllable has been defined as a “unit at a higher level than that of the phoneme or sound segment, yet distinct from that of the word or morpheme” (Gimson 1980: 55–57). Two types of approaches have been adopted to define the term: (1) a phonetic approach, which tries to provide a universal definition of the concept; and (2) a linguistic or functional approach, which makes reference to the structure of a particular language. In the phonetic approach, syllables are generally described as consisting of a centre (nucleus or peak), which involves little or no obstruction to the airflow and is perceived as comparatively loud (Roach 2005: 67). This nuclear place is generally realised by vowels although in some circumstances it may also be filled by a consonant, as in the third syllable of the word eleven [ɪˈlevn̩] (Section 1.3.3.2).6 Minimum syllables consist only of a nucleus (e.g. /ɑː/ are, /ɔː/ or). In addition, before or after the nucleus, that is, at the beginning (onset) or end (coda) of the syllable, there may also be greater obstruction to the airflow and/ or less sound. Syllables may thus consist of an onset, followed by a nucleus and a coda (e.g. /gʊd/ good), or else they may just consist of either an onset and a nucleus (e.g. /tiː/ tea) or a nucleus and a coda (e.g. /ɪz/ is). The notion of rhyme, comprising the peak (or non-consonantal segment(s)) and the coda (the final consonantal segment(s)), is used to postulate a close relationship between these two elements that is distinct from the onset (initial consonantal segment(s)) because, among other things, word stress is assigned considering only the elements of the rhyme (rhyme projection principle), as shown in Figure 8 which plots the relative prominence and boundaries of the three syllables in the word forbidden.

6 Sounds are usually represented along a “sonority hierarchy” depending on their “carrying power” or relative sonority: at the top are open vowels, followed by close vowels, laterals, nasals, approximants, trills, fricatives, affricates and plosives, and flaps at the end; and, generally, voiced sounds are more prominent than voiceless ones (Cruttenden 2014) (see § 1.3.3.3).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

20

Phonetics and Phonology

Figure 8: Plotting syllable prominence and syllable boundaries

AI 1.3 Identification of a short vowel as syllabic nucleus In the linguistic (functional or phonological) approach, syllables are examined taking into account the possible combinations of phonemes that occur in a given language. Syllabic structure differs from one language to another in terms of such parameters as how many consonants can occur in the onsets and codas of syllables, or whether vowels can occur without consonantal onsets (syllables with zero onsets), or whether both open syllables (those ending in a vowel) and closed or checked syllables (those ending in a vowel) are possible.7 As an illustration, the three English words /tʌk/ tuck, /trʌk/ truck, and /strʌk/ struck are all structurally different. Using C to mean ‘a consonant’ and V to mean ‘a vowel’ (a diphthong is considered as one unit and is thus represented as a single V), structural formulae for these three words are CVC, CCVC and CCCVC, respectively. Series of two Cs or more indicate the presence of a consonant cluster, that is, a group of consonant sounds that appear together without an intervening vowel in any possible position in a word. Formulae of this sort are usually used to comment on the phonological structure of individual syllables. If we go back now to the notion of minimal pair, in the light of syllable structure, we could say that, while /pæn/ pan and /tæn/ tan, both of CVC structure, form a minimal pair in English, /pæn/ pan and /tæp/ tap do not because, although both are also of CVC structure, they differ in more than one structural place. The words /spæn/ span and /tæn/ tan do not form a minimal pair either, because their structures (CCVC and CVC) are not identical. 1.3.3.2 Syllabic patterns in English and Spanish and advice Now let us compare the syllabic patterns of English and Spanish outlined in Table 2 below (adapted from Gallardo del Puerto 2005: 208–225; see also Gimson 1984: 237; Roach 2005: 73–77; Cruttenden 2014: 259–267). 7 Checked vowels are those that occur in checked or closed syllables (Crystal 2008: 74).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

21

Table 2: Syllable patterns in English (RP) and Spanish (PSp) PATTERNS

ENGLISH

SPANISH

Open syllables 1 2 3 4

V CV CCV CCCV

/eɪ/a /peɪ/ pay /preɪ/ pray /spreɪ/ spray

/i/ y /mi/ mi /tri/ triturar

Closed syllables 1 2 3 4 5 6 7 8 9 10 11 12 13 14

VC VCC VCCC CVC CVCC CVCCC CVCCCC CCVC CCVCC CCVCCC CCCVC CCCVCC CCCVCCC CCCVCCCC

Initial clusters

CC

/tw/ /kw/ /pj/ /tj/ /kj/ /pr/ /br/ /tr/ /dr/ /kr/ /gr/ /pl/ /bl/ /kl/ /gl/ /fr/ /θr/ /ʃr/ /fl/ /sp/ /st/ /sk/

/iːt/ eat /iːts/ eats /ɑːskt/ asked /biːt/ beat /biːts/ beats /θæŋks/ thanks /teksts/ texts /kwiːn/ queen /kwiːnz/ queens /trʌŋks/ trunks /striːt/ street /striːts/ streets /stræŋgl ̩/ strangle /stræŋgld̩/ strangled

/ar/ árbol /eks/ extra

tweed question pure tube cure prime brie tree draw crew green plant blue clue glue free three shrink flee spy stay sky

tuétano cueva pié tiene quiere preso brasa triturar dragón cruz gramo pleno blusa cloro glucosa Francia X X flojo X X X

/sol/ sol /teks/ texto

/tren/ tren /trans/ transcribir

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

22

Phonetics and Phonology

Table 2: (continued) PATTERNS

ENGLISH

SPANISH

/sf/ /sm/ /sn/ /sl/ /sw/ /sj/

sphere smog snake slay swim assume

X X X X suave siento

Initial clusters

CCC

/spl/ /spr/ /str/ /skr/ /spj/ /stj/ /skw/ /skj/

splash spray street screen spurious stew squid skew

X X X X X X X X

Final clusters

CC

/pt/ /bd/ /kt/ /gd/ /θt/ /ðd/ /ft/ /vd/ /st/ /zd/ /ʃt/ /ʒd/ /sk/ /sp/ /tʃt/ /dʒd/ /mp/ /nt/ /ŋk/ /nd/ /md/ /ŋd/ /lt/ /lp/ /lk/ /tθ/ /dθ/ /ps/

opt described sect begged earthed breathed soft curved fast raised washed camouflaged ask wasp watched judged camp ant think and seemed longed belt help milk eighth width cops

X X X X X X X X X X X X X X X X X X X X X X X X X X X X

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

Table 2: (continued) PATTERNS

Final clusters

CCC

ENGLISH

SPANISH

/bz/ /bs/ /ts/ /dz/ /ds/ /ks/ /ɡz/ /fθ/ /fs/ /vz/ /θs/ /ðz/ /nθ/ /ŋθ/ /mf/ /ns/ /nz/ /mz/ /ŋz/ /lf/ /lv/ /ls/ /lz/ /lθ/ /lʃ/ /ntʃ/ /ndʒ/ /tn/ /dn/ /lm/ /tl/ /dl/ /rs/

jobs X cuts beds X sex legs fifth cliffs halves baths bathes tenth length triumph sense cans comes songs golf twelve else tells wealth Welsh branch change button garden elm battle pedal X

X substraer X X adscrito exponer X X X X X X Sánz X X instinto X X X golf X vals X X X X X X X X X X perspicacia

/spt/ /skt/ /mpt/ /ŋkt/ /lpt/ /lkt/ /ntʃt/ /ndʒd/ /ltʃt /ldʒd/

grasped asked bumped linked helped milked launched changed filched divulged

X X X X X X X X X X

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

23

24

Phonetics and Phonology

Table 2: (continued) PATTERNS /pst/ /tst/ /dst/ /kst/ /mft/ /lvd/ /lst/ /dnt/ /znt/ /vnt/ /pts/ /kts/ /sps/ /sts/ /sks/ /fts/ /mps/ /nts/ /ndz/ /ŋks/ /lps/ /lbz/ /lts/ /ldz/ /lks/ /pθs/ /tθs/ /dθs/ /ksθ/ /fθs/ /mfs/ /nθs/ /lfs/ /lvz/ /lθs/ /lfθ/ /lmz/

ENGLISH

SPANISH

eclipsed mightst midst text triumphed involved whilst hadn’t hasn’t haven’t opts products grasps atheists masks lifts camps ants winds thanks helps bulbs belts worlds silks depths eighths widths sixth fifths triumphs months Ralph’s wolves wealths twelfth films

X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X X

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

25

In Table 2 you can see that neither English nor Spanish exploit all the possible combinations of phonemes within the syllable because they are not allowed by the phonological patterns (or phonotactics) of these two languages. It can also be observed that English has a wider range of syllable patterns than Spanish. As will be explained, besides three-consonant codas, English also admits the possibility of four-consonant codas generally resulting from the addition of tense and/or number endings to three-consonant codas (glimpsed /ɡlɪmpst/, twelfths /twelfθs/). Neither coda type is found in Spanish and twoconsonant codas are very infrequent (Gallardo del Puerto 2005: 208–225). English permits open syllables with a maximum of three consonants in their onsets, and closed syllables with up to four consonants in their codas, which makes at least eighteen syllabic patterns: four onset and fourteen coda patterns (Gimson 1984: 237; Roach 2005: 73–77; Cruttenden 2014: 259–267). The four onset patterns that may occur in open syllables are listed as patterns 1 to 4 in Table 2. Now some specifications should be made about patterns 3 and 4. In pattern 3, CCV (/pleɪ/ play), CC is either /s/ + /t w m/ (/steɪ/ stay, /swæm/ swam, /smæʃ/ smash) or C + /l r w j/ (/kleɪ/ clay, /traɪ/ try, /kwɪk/ quick, /ˈfjuːəl/ fuel). This suggests that that initial two-consonant clusters may follow one of two possible patterns: /s/ + /t w m/ or C (one of a set of fourteen consonants) + /l r w j/. Turning to pattern 4, CCCV = /s/ + /p t k/ + /l r j w/ (/sprei/ spray, /stjuːd/ stewed, /skwɔːk/ squawk, /skwiːl/ squeal). This specification indicates that onsets with three-consonant clusters have to consist of initial /s/, followed by one element of the /p t k/ set, and a third one from /r l j w/. Note that there is a constraint to the effect that when the third consonant is /w/, the first two consonants must be /s/ and /k/ (/skwɔːk/ squawk, /skwiːl/ squeal). Further combinatorial constraints on CCCV patterns determine that /spr/ and /str/ are both permitted, but /spw/ and /stw/ are not. Turning to closed syllables, Table 2 above lists the fourteen coda patterns that occur in English. However, there exist a number of restrictions regarding the types of consonants that may be combined in closed syllables, the most important ones being: (1) /r h j w/ do not combine with other consonants in final positions. (2) CC coda clusters result from the following combinations: /p t k/ + /t s θ/ (/biːts/ beats, /striːts/ streets, /iːts/ eats)), /b d g/ + /d z/ (/begd/ begged), /tʃ/ + /t/ (/bliːtʃt/ bleached), /ʤ/ + /d/ (/brɪʤd/ bridged), nasal + plosive, fricative or affricate (/kwiːnz/ queens), fricative + plosive (/briːðd/ breathed), /l/ + stops (plosives or nasals) or fricatives (/kɔːld/ called). (3) CCC coda clusters predominantly follow short vowels and result from three main patterns: /p t d k/ + /ts st θs sθ/ (/siksθ/ sixth), nasal or /l/ + a plosiveplosive, plosive-fricative, fricative/affricate-plosive (/θæŋks/ thanks, /trʌŋks/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

26

Phonetics and Phonology

trunks, /stræŋgl ̩/ strangle), and /s/ + a plosive-plosive or a plosive-fricative combinations (ɑːskt/ asked). (4) CCCC coda clusters occur only rarely and are the result of attaching /t/ or /s/ morphemes to CCC (/teksts/ texts, /siksθs/ sixths, /ɡlɪmpst/ glimpsed). Recapitulating what has been just described, Figure 9 below represents the structure of the syllable in English. It shows that slots 1, 2 and 6 tend to be occupied by “pure” consonants, whereas slots 3 and 5 are filled by one approximant /l r w j/ (see Sections 2.3.2.3 and 4.2.4) or the homorganic nasal phoneme /n/; while slot 4 is occupied by a vocalic element (or a sequence of them), and positions 7 and 8 are usually realised by morphological endings (e.g. /s z/, /t d/) (Knowles 1974).

Figure 9: The structure of the syllable in English

Spanish, on the other hand, permits open and closed syllables with up to two consonants in their onsets and codas, which adds to at least nine possible syllable patterns: three onset and six coda patterns, which are illustrated in Table 2 and are reproduced below for clarity of exposition as (1) to (3) and (4) to (9) below, respectively (Quilis 1993; Quilis and Fernández 1996): (1) V (/i/ y ‘and’, /oi/ hoy ‘today’) (2) CV (/mi/ mi ‘my’, /su/ su ‘his, her’) (3) CCV (/fru-/ as in triturar ‘grind’, fruta ‘fruit’) (4) VC (/ar/ as in arbol ‘tree’); (5) VCC (/eks/ as in extra ‘extra’); (6) CVC (/sol/ sol ‘sun’); (7) CVCC (/teks/ as in texto ‘text’); (8) CCVC (/tren/ tren ‘train’); (9) CCVCC (/trans/ as in transcribir ‘transcribe’). Taking the above into account we can conclude that in English there is a predominance of closed syllables and there exists a wider variety of syllable structures (17 vs. 9), favouring consonantal clusters in both onsets and codas

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

27

that are not allowed in Spanish. In Spanish, on the other hand, open syllables are strongly favoured and syllables ending with consonants are less common, the dominant syllable pattern being CV, followed by CVC. Onset and coda clusters have a maximum of two consonants, words having two consonants in coda position being mostly of foreign origin (duplex). Spanish learners are thus advised to pay particular attention to the eight English syllabic patterns that have no correlate in Spanish: CCCV, VCCC, CVCCC, CVCCCC, CCVCCC, CCCVC, CCCVCC and CCCVCCC. Yet another difference between English and Spanish is that, while in Spanish the V element of the syllable can only be a vowel, in English it can quite frequently be a sonorant consonant /l r m n ŋ/, which therefore becomes syllabic, as is explained in Section 1.3.3.3. Consonant clusters may be problematic for SSLE (Sedláčková 2010; Gallardo del Puerto 2005: 316–322; 356–358, 373–375, 407–427, 437–450; Estebas Vilaplana 2009: 63, 98–117). In principle, initial consonant clusters that are shared by both languages should pose little or no difficulty: /pl/ (plant – planta), /pr/ (promise – promesa), /bl/ (blouse – blusa), /br/ (broccoli – brócoli), /tr/ (trace – traza), /dr/ (drug – droga), /kl/ (class – clase), /kr/ (crisis – crisis), /gl/ (glory – gloria), /gr/ (great – grande), /fl/ ( flower – flor) and /fr/ ( frank – franco/a). If the same consonant appears twice in spelling, the pronunciation only involves one sound (letter /ˈletə/). There are two exceptions to this tendency. One is when the same affricate is pronounced twice in a sequence, as in rich church /rɪtʃ tʃɜːtʃ/, orange jam /ˈɒrɪndʒ dʒæm/; and the other is when the two consonants belong to different grammatical categories, as in midday /ˌmɪdˈdeɪ/ (prefix + root), drunkenness /ˈdrʌŋkənnəs/ (root + suffix), part-time /ˌpɑːt ˈtaɪm/ (compound), late train /ˈleɪt ˈtreɪn/, in which case double sounds are pronounced like single sounds but with longer duration, even though two sounds appear in transcriptions. More troublesome seem to be those consonant clusters that exist in English but do not occur in Spanish, in which case SSLE usually opt for either cluster reduction (simplification) or the substitution or adjustment of the sound that is alien to them so as to come “closer” to Spanish pronunciation. In the former case frequently the final consonant(s), generally /t/, /d/, /k/ or /s/, are omitted so that such words as texts /teksts/, talked /tɔːkt/ and complained /kəmˈpleɪnd/ may be mispronounced as *[teks], *[tɔːk] and *[kəmˈpleɪn] and the listener will understand a singular referent and present actions rather than plural referents and past actions, respectively. But there may be cases in which the middle consonant (child’s *[tʃaɪlz], nine years *[naɪjɪǝz]) or the last two consonants (holds *[xol]) are dropped. Turning to adjustments, the most common one is that of initial English clusters beginning with /s/, not found in Spanish, which tend to be wrongly pronounced with a preceding epenthetic, or “added”, /e/ sound. As a result, such words as

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

28

Phonetics and Phonology

star and smart may be mispronounced as *[esˈtɑɾ] and *[esmɑɾt], instead of /stɑː/ and /smɑːt/ (Carlisle 1998; Sedláčková 2010). Otherwise, a vowel, usually [e], may also be inserted between the two final consonants, and so words such as strangle /stræŋglŋ/ and talked /tɔːlkt/ may be wrongly pronounced as *[ˈstræŋgel] and *[ˈtɔːlked], causing a strong foreign accent effect because either a pronunciation with [ə] or preferably one with no vowel between the consonants would sound closer to native. Other phenomena resulting from universal co-articulatory tendencies may also occur such as devocing of word-final sounds (climbed *[klaɪmt], love *[lʌf], years *[jɪǝs], matches *[mætʃɪs], evenings *[iːvnɪŋks], mornings *[mɔːnɪŋks], pads *[pæts], always *[ɔːlweɪs], lose *[luːs]), fricativisation of word-final plosives (helped *[xelps]), and metathesis, or the alteration of the order of phonemes in a word (last *[lats]). SSLE should be aware of these mispronunciations and should try to pronounce English consonant clusters appropriately so as to avoid communication and intelligibility problems. This is particularly relevant for speakers of some Spanish accents, such as the Andalusian accent, in which word-final consonants, especially [s], tend to be dropped.

1.3.3.3 Syllabic consonants Syllabic consonants are widely produced in various registers in English (Jones 1967/1976; Roach, Sergeant and Miller 1992) but they do not occur in Spanish (Finch and Ortiz Lira 1982). Marked with a short vertical stroke beneath the consonant symbol [n̩], syllabic consonants frequently occur in CC̩ syllabic patterns that generally result from the weakening, or reduction, of CəC syllables (eleven [ɪˈlevn̩]). Such syllabics are pronounced as the centre of the syllable instead of a vowel, after a preceding consonant in post-tonic unstressed positions and almost always just before a morpheme boundary, as in, for instance, [ˈbɒtɫ ̩ŋ] bottle and [ˈbɒtɫ ̩ŋɪŋ] bottling. In RP there are four consonants that can be syllabic: [ɫ ̩] /l ̩/, perhaps the most noticeable example, and [n̩ m̩ ŋ̩]. /r/ may also be syllabic in rhotic accents,8 but in RP it becomes syllabic only in some sporadic cases, generally as a result of /ə/ deletion in rapid speech which may leave either syllabic [ɹ̩] or pre-vocalic non-syllabic /r/ pronunciations (generous [ˈdʒenəɹəs] → [ˈdʒenɹ̩əs] – [ˈdʒenɹəs], memory [ˈmeməri] → [ˈmemɹ̩i] – [ˈmemɹi]) (Roach 2005: 86–90, 125–127). 8 Rhotic accents, languages and dialects are those in which /r/ is pronounced following a vowel (such as Spanish, as in par /par/ ‘pair’). In contrast, varieties which do not have this feature are non-rhotic (such as RP as in par /pɑː/ and park /pɑːk/) (Crystal 2008: 417).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonology

29

Concerning syllabic [ɫ ̩] /l ̩/, it produces laterally released preceding consonants (with lowered side-rims of the tongue allowing the airflow to escape over them) and is found word-medially and word-finally, corresponding to the following spellings: (petal [ˈpetɫ ̩]), (panel [ˈpænɫ ̩]), (e.g. bottle [ˈbɒtɫ ̩], couple [ˈkʌpɫ ̩] struggle [ˈstrʌgɫ ̩], cattle [ˈkætɫ̩]) and (mostly in plural nouns and third personal singular present verbal forms, as in e.g. wrestles [ˈresɫ ̩z]). In words with final in their base forms, if a suffix beginning with a vowel is attached, then the [l] generally remains syllabic but the letter is lost (e.g. struggling [ˈstrʌgŋɫ ̩ɪŋ] and bottling [ˈbɒtɫ̩ɪŋ]). On the other hand, similar words that are not formed in this way do not have syllabic [ɫ ̩] and therefore reflect a grammatical contrast such as the one existing between [ˈkɒdɫ ̩ɪŋ] coddling, participial form of the verb coddle, and [ˈkɒdlɪŋ] codling meaning ‘small cod’. Turning to the nasal group, syllabic [n̩ ] is the most frequent of all. It usually occurs word-medially and word-finally after alveolar plosives and fricatives, which are nasally released by lowering the soft palate, as in threaten [ˈθretn̩], threatening [ˈθretn̩ɪŋ], seven [ˈsevn̩], often [ˈɒfn̩], heaven [ˈhevn̩]. But [n̩] does not occur after [l tʃ ʤ], and so such words as sullen, Christian, and pigeon, for example, must be pronounced [ˈsʌlən], [ˈkrɪstʃən], and [ˈpɪʤən]. Similarly, after velar consonants in syllables spelt -an or -on, [n̩] is rarely heard (e.g. toboggan [təˈbɒɡən], wagon [ˈwæɡən]), while after bilabial consonants [n̩] and [ən] pronunciations alternate (happen [ˈhæpn̩], [ˈhæpən], happening [ˈhæpn̩ɪŋ], [ˈhæpənɪŋ]). Lastly, syllabic [m̩ ŋ̍ ] usually result from progressive assimilatory and elision processes as it occurs when preceding a bilabial (ribbon [ˈɹɪbm̩ ], happen [ˈhæpm̩ ]) or a velar plosive (thicken [ˈθɪkŋ̍], dragon [ˈdræɡŋ̍]), as will be further explained in Section 5.2.7.

AI 1.4 Syllabic consonants in RP It should be kept in mind that, although every CC̩̩ has a corresponding CəC alternative, not every CəC has a corresponding CC̩̩ counterpart. These alternations can be explained by resorting a phonological rule that is related to the sonority hierarchy principle (Jones 1975), illustrated in Figure 10:

Figure 10: Sonority scale or sonority hierarchy

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

30

Phonetics and Phonology

Although there is some argument about its details, the main tenets of the sonority hierarchy principle are widely accepted. Based on the principle that the more open the VT, the more sonorous or prominent the speech sound is, the sonority hierarchy suggests that consonants are perceived as less sonorous than vowels, and that plosives are the less sonorous of speech sounds as opposed to open vowels, which are the most sonorous speech sounds, while the rest show increasing degrees of sonority from left to right in the sonority hierarchy. Applying the sonority hierarchy principle to English syllabicity the following phonological rule emerges: Cx ə Cy (V) → Cx C̩ y (V) #, under the conditions that (a) C̩ is post-tonic, (b) Cy = non-glide sonorant consonant, (c) Cx < Cy (where “ Cy, which contravenes the syllabic sonorant creation principle barring syllabic consonant pronunciations: venom [ˈvenəm], melon [ˈmelən], enemy [ˈenəmi], felony [ˈfeləni]. In semiformal registers of RP, syllabic consonants are often regarded as obligatory, in so far as non-syllabic pronunciations may be considered as mispronunciations: for example, saying [ˈpiːpəɫ] rather than [ˈpiːpɫ ̩] for people, or [ˈkætəɫ] rather than [ˈkætɫ ̩] for cattle. In more technical and less common words, however, syllabic and non-syllabic pronunciations oscillate more freely, although syllabics are usually preferred (e.g. missal [ˈmɪsɫ̩], [ˈmɪsəɫ], acquittal [əˈkwɪtɫ ̩], [əˈkwɪtəɫ]).9 In pronunciation dictionaries (LPD, EPD), alternative pronunciations with or without syllabic consonants are transcribed with a superscript schwa [ə], as in the case of awful [ˈɔːfəɫ], while syllabics are marked when that is the preferred pronunciation (for further details on /l/ pronunciations see § 4.2.5.1). What the above examples of syllabics have in common is that all of them result in unstressed two-consonant clusters, where the syllabic one is the second element. This seems to suggest that three-consonant clusters with a third (or second and third) syllabic element are disfavoured in RP, although they may 9 Note that the word-final ending ( fertile, futile, missile, reptile) is pronounced [aɪɫ] in RP, but in American English is reduced to syllabic dark [ɫ ̩].

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonetic transcription

31

also be possible in some words (including those with two syllabics in a row, [CC̩C̩], as in e.g. national [ˈnæʃn̩ɫ ̩], veteran [ˈvetr̩n̩]), and optional [ˈɒpʃn̩ɫ ̩]). Instead, alternative pronunciations are preferred such as: (1) [CCəC] (e.g. abandon [əˈbændən], Boston [ˈbɒstən], Camden [ˈkæmdən], lantern [ˈlæntən], London [ˈlʌndən], veteran [ˈvetrən]), optional [ˈɒpʃnəɫ]; (2) [CəCC̩] (e.g. [ˈvetərn̩]); (3) [CəCəCə] (e.g. [ˈvetərən]).

1.4 Phonetic transcription One of the chief resources of phonetics is phonetic transcription, where the term “phonetic” is used in a broad sense to refer to the representation of speech using phonetic symbols. In what follows we shall see the phonetic symbols that are used to represent sounds as presented in the International Phonetic Alphabet (section 1.4.1); the two main types of phonetic transcription that are used to represent speech sounds, broad and narrow transcription (section 1.4.2); and some practical guidelines to help you transcribe better (section 1.4.3).

1.4.1 The International Phonetic Alphabet (IPA) The International Phonetic Alphabet (IPA)10 (Table 3 below) is the best-known and most widely accepted set of phonetic symbols that are used to classify speech sounds. In addition, the IPA provides many diacritics, which are small additional symbols that are added to the main symbol shapes to modify the sound being represented. Diacritics may be of two kinds: (1) those representing suprasegmental information (primary and secondary stress marks and so on), and (2) those representing different articulatory features (e.g. tongue advancement and retraction, nasalisations, degrees of lip rounding and voicing, among many others), which depending on the shape of the phoneme symbol may be placed either above or below it and which are necessary to represent the degree of allophonic detail that is provided in narrow transcriptions, as will be explained in Section 1.4.2 and Chapter 5.

10 The acronym IPA is used to refer to both the association and to its alphabet.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

32

Phonetics and Phonology

Table 3: The symbols of the IPA

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Phonetic transcription

33

AI 1.5 Becoming familiar with the IPA Phoneticians use a larger set of symbols when they want to represent sounds more accurately, and/or including non-speech sounds. Table 3 shows that in the IPA, cardinal vowels are included in a vowel quadrilateral chart (Section 2.3.1.3), while consonants are arranged in a table according to place of articulation, manner of articulation and voicing (Section 2.3.2).

1.4.2 Types of phonetic transcription Transcription can be in either of two “directions”. From what we hear spoken, or what we read from “spelt” text (i.e. in ordinary writing), we can write a so-called direct transcription into phonemic or phonetic text, while from phonetic symbols we can write a reversed transcription into spelt text. Direct is more often needed than indirect. We must also be clear about another distinction (that has just been mentioned in passing). Very often we do not need to indicate the pronunciation of a word very precisely, but just enough to distinguish it, for example, from other words that are fairly similar but perhaps spelt the same in ordinary writing. Tear is pronounced /tɪə/ when referring to a teardrop that comes from your eye (Sp. lágrima) but /teə/ when meaning ‘to damage something by pulling it apart or into pieces or by cutting it on something sharp’ (OLD) (Sp. rasgar). Such basic transcriptions are variously called broad, phonemic or phonological and show very little phonetic detail and especially not detail which is predictable from general phonetic principles or from the phonological properties of the language in question (SID). Thus, aspiration of plosives or nasalisation of vowels would not be shown in a broad transcription of English, but would be shown in languages where these features are phonologically contrastive, i.e. when they imply a change of meaning. In broad transcriptions we work with phonemes and they are enclosed between slant bars (e.g. ten /ten/) (Sections 1.3.1 and 1.5). The number, identity and membership of phonemes differs from language to language, and from accent to accent within a given language. Phonemic transcriptions therefore record the order in which phonemes occur using a set of symbols and a series of conventions and principles which must be studied in order to assign each symbol its correct value, the basic operating principle being that of using one symbol for phoneme. Thus, in Section 1.5. it will be shown that to produce a phonemic transcription of RP, for instance, forty-four symbols will be used: twenty for vowels and vowel glides, and twenty-four for consonants.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

34

Phonetics and Phonology

Narrow transcription, on the other hand, shows a large amount of phonetic detail (even non-contrastive), using additional diacritics to give extra information about the realisation of each sound in a particular context (such as nasalisations, (de)voicings, dentalisations, and length reductions), and it is enclosed between square brackets (e.g. ten [tʰẽn̊]). For this reason this kind of transcription is also called phonetic or allophonic. The latter label derives from allophone, meaning that in narrow transcriptions we include information about the allophonic realisations of phonemes (see section 1.3.1. and Chapters 3 to 5). Allophonic transcriptions therefore record the actual realisations of each phoneme in its precise context of occurrence, employing a set of phonetic symbols and diacritics. For ease and clarity of exposition, except for Chapter 5, we mostly show just the phonetic (or allophonic) detail of one segment in an otherwise phonemic transcription. These transcriptions are enclosed between square brackets ([tʰen]), because slant bars are restricted to phonemic transcriptions (/ten/). In the past, broad and narrow transcriptions were identified with phonemic and allophonic transcriptions, respectively, but nowadays, while the tendency is still for phonemic transcription to be always broad, phonetic transcriptions may be relatively broad or narrow, depending on the amount of phonetic detail provided. All types of transcriptions are useful for phonetic analysis. A phonemic transcription is more economical and it is the one given in pronunciation dictionaries. But mastery in allophonic transcription is also essential for both foreign language teachers and learners, not only to refer to the actual nature of the target sounds which in many cases involve very subtle differences, but also to avoid mispronunciations (e.g. tree should be pronounced [ṯɹ̥iːː] not *[tɾ̥iː]) (Chapters 4 and 5). In this book, however, targeting mostly phonemic contrasts, emphasis is laid on broad (direct or reversed) transcription as it includes all (and only) the information that is necessary to meaning or otherwise that is most relevant in one particular context. Narrow transcriptions are also used in Chapters 3 to 5 in order to explain the variations of RP phonemes that are contextually determined and the mispronunciations of SSLE.

AI 1.6 An example of broad transcription An example of broad and narrow transcriptions Consider the examples of broad and narrow transcriptions presented below, which are preceded by their orthographic version. Look at the text, trying to match the written words and the symbols. Although in terms of methods they are certainly very different from the rest, phonetics and phonology can be regarded as two further interrelated internal

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Received Pronunciation (RP) and Peninsular Spanish (PSp)

35

fields of linguistics in that, broadly, both are concerned with the same object of study: the phonic or sound component of language, an indispensable foundation of linguistic enquiry. || ɔːlˈðəʊ ɪn tɜːmz əv ˈmeθədz ˈðeɪ ə ˈsɜːtnli ˈveri ˈdɪfrənt frəm ðə rest | fəˈnetɪks ənd fəˈnɒləʤi kən bi rɪˈɡɑːdɪd əz tu: ˈfɜːðər ˌɪntərɪˈleɪtɪd ɪnˈtɜːnl ̩ fiːldz əv lɪŋˈɡwɪstɪks ɪn ðæt | ˈbrɔːdli | bəʊθ ə kənˈsɜːnd wɪð ðə seɪm ˈɒbʤekt əv ˈstʌdi | ðə ˈfɒnɪk ɔː ˈsaʊnd kəmˈpəʊnənt əv ˈlæŋɡwɪʤ | ən ˌɪndɪˈspensəbl ̩ faʊnˈdeɪʃn̩ əv lɪŋˈɡwɪstɪk ɪnˈkwaɪəri || A phone is a single phonetic segment viewed in terms of its phonetic character without regard to its possible phonological status. The term ‘phone’ is related to ‘phoneme’ in the same way ‘morph’ is related to ‘morpheme’. || ə fəːʊ̃ n ɪz ə ˈsʲɪ ̃ŋɡˡɫ ̩ fə̃ˈnet̬ʲ ɪk ˈseɡⁿmə̃nt vʲjuːd ɪ ̃n tʰɜ̃ːmz əv ɪts fə̃ˈnet̬ʲ ɪk ˈkæɹək˺tə wɪˈðaˑʊt ɹʲɪˈɡɑːd tʰə ɪts ˈp̠ʰʷɒs̬ʲ ɪbˡɫ ̩ ˌfəːʊ̃ nəˈl ̠ʷɒʤʲɪk̬ˡɫ ̩ ˈsteˑɪt̬əs || ðə tʰɜ̃ːm fəːʊ̃ n ɪz ɹʲɪˈleˑɪt̬əd tʰə ˈfəːʊ̃ n ʲĩːm ɪ ̃n̪ ðə s̬eːɪ ̃m weːːɪ m̠ ʷɔːf ɪz ɹʲɪˈleˑɪt̬əd tʰə ˈm̠ ʷɔːf̬ʲ ĩːm ||

1.5 Received Pronunciation (RP) and Peninsular Spanish (PSp) This manual describes Received Pronunciation (RP for short), also known as BBC English, Oxford English, or the Queen’s English (Jones 1967/1976; Gimson 1980, 1984, 1994; Wells 1982). A widely established term, RP is endorsed here to refer to the standard prestige variety of British English that is “received” through the mass media, particularly the BBC (but it is not obligatory), and socially, it is associated with the upper and upper-middle class (public school education, stockbrokers, diplomats, etc.). RP originated in the Southeast, and probably it is still associated with this area by people from Northern, Central and Western England.11 There are two main reasons to choose RP as the model accent: (1) It is a neutral accent that is universally understood by both native and foreign speakers alike; and (2) It is the accent of reference that is described in most books and materials of EFL, especially in Europe, and so it should also be familiar to Spanish learners of English.12 11 RP does not designate a monolithic accent, but rather is an umbrella term that is used to refer to a number of varieties such as: U-RP (upper-crust RP) (old-fashioned upper class accent, almost a caricature), Conservative RP (characteristic of older generations), Mainstream RP (“regular” accent of native RP speakers), Adoptive RP (spoken by adults who did not speak RP as children) or Near-RP (an accent that does not adjust itself to the definition of RP because it includes some regionalisms) (Wells 1982; Hughes et al. 2013). 12 Other manuals prefer to speak about Standard Southern British English (SSBE) (Mott 2005) in compliance with the recommendation of the Handbook of the International Phonetic Association (1999: 4) that is best to use the label “standard”, whereas such terms as General

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

36

Phonetics and Phonology

Table 4 below shows that the sound system of RP consists of forty-nine sounds, twelve vowels, eight diphthongs, five diphthong + /ə/ combinations and twenty-four consonants (Cruttenden 2014; Roach 2005), whereas the sound inventory of PSp has forty-two sounds, five vowels, fourteen diphthongs, four triphthongs and nineteen consonants (Navarro Tomás 1966 [1946], 1991 [1918]). Now listen to all the sounds of English RP and Peninsular Spanish displayed in Table 4.

AI 1.7 The sound systems of RP and PSp The layout of Table 4 is intended to show phoneme correspondences or closest sound equivalences between RP and PSp. Accordingly, blank lines are used when no exact matchings occur (as in the case of RP /ŋ/ that has no equivalent phoneme in Spanish), while the use of different phonemic symbols across the two languages shows that the sound(s) in question exhibit some phonetic similarity, but not close phonetic likeness (as in the case of RP /h/ and PSp /x/, which are both fricative sounds but the former is glottal while the latter is velar) (Gallardo del Puerto 2005). A detailed description of the phoneme inventory of RP in comparison with that of PSp is provided in Chapters 3 (examining vowels, diphthongs and triphthongs individually) and 4 (consonants). At this early stage it should only be noted that our comparison is not systematic, but rather it focuses on the differences and parallelisms that exist between the sound inventories of these two languages that are felt useful to better grasp English phonetics and pronunciation by speakers of Spanish. Consequently, no individual description is offered for those phonemes that occur only in Spanish (e.g. [ʎ] or [ɲ] in llama ‘flame’ or caña ‘rode’, respectively), but observations about allophonic and phonemic realisations in varieties of Spanish and the languages spoken in Spain are included whenever necessary to clarify the discussion or aid learners to gain pronunciation strategies and skills. British (GB) (Cruttenden 2014: 4–5) and non-regional pronunciation (NRP) (Collins and Mees 2009: 4–6) are also employed to refer to an allegedly more updated and more encompassing neutral type of modern English with no local accent features. In addition, it has also been claimed that Standard Scottish English (SSE) is an easier accent for foreign learners than RP on the grounds that it is a rhotic accent (Cruttenden 2014: 326), while other models are used to represent the target pronunciation in situations in which English is used as a lingua franca, as in the case of Amalgam English (which incorporates the more easily learnable characteristics of various Englishes and some features that are common to subcontinental varieties, e.g. Bangladesh, India and Pakistan) and International English (which reduces the inventory of English sounds even more to what is felt more easily learnable). Finally, those countries that have been traditionally influenced by the U.S. tend to take General American (GA) as a model.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Received Pronunciation (RP) and Peninsular Spanish (PSp)

Table 4: The sound systems of RP and PSp English Sounds Examples (49)

Spanish Sounds Examples (42)

Vowels (12)

Vowels (5)

æ ʌ ɑː e ə ɜː ɪ iː ɒ ɔː ʊ uː

bad [bæd] cut [kʌt] bard [bɑːd] bed [bed] colour [ˈkʌlə] herb [hɜːb] sit [sɪt] seat [siːt] cod [kɒd] cord [kɔːd] pull [pʊl] pool [puːl]

a

cama [ˈkama] ‘bed’

e

mesa [ˈmesa] ‘table’

i

sin [sin] ‘without’

o

amor [aˈmor] ‘love’

u

uso [ˈuso] ‘usage’

Dipthongs (8) aɪ eɪ ɔɪ aʊ əʊ

eə ɪə

ʊə

life [laɪf] grey [greɪ] boil [bɔɪl] cow [kaʊ] so [səʊ]

Diphthongs (14) ai ̯ ei ̯ oi ̯ au̯

aire [ˈai ̯re] ‘air’ rey [rei ]̯ ‘king’ hoy [oi ]̯ ‘today’ pausa [ˈpau̯sa] ‘pause’

eu̯ ou̯

neutro [ˈneu̯tɾo] ‘neutral’ bou [bou̯] ‘type of fishing’

ja je jo ju

viaje [ˈbjaxe] ‘trip’ diente [ˈdjen̻te] ‘tooth’ radio [ˈraðjo] ‘radio’ viuda [ˈbjuða] ‘widow’

wa we wo wi

cuadro [ˈkwaðɾo] ‘painting’ bueno [ˈbweno] ‘good’ asiduo [aˈsidwo] ‘assiduous’ fuimos [ˈfwimos] ‘(We) went’

bear [beə] beer [bɪə]

sure [ʃʊə]

Diphthongs + [ə] / Triphthongs (5) eɪə aɪə ɔɪə əʊə aʊə

player [pleɪə] tyre [taɪə] employer [ɪmˈplɔɪə] slower [sləʊə] hour [aʊə]

Diphthongs + [ə] / Triphthongs (4)

wei ̯ wai ̯ jei ̯ jai ̯

buey [bwei ]̯ ‘ox’ averigüáis [aβ̞eɾiˈγwai ̯s] ‘(You) find out’ vieira [ˈbjei ̯ɾa] ‘scallop’ saciáis [saˈθjai ̯s] ‘(You) satisfy’

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

37

38

Phonetics and Phonology

Table 4: (continued) Consonants (24) p b t d k ɡ f v θ ð s z ʃ ʒ h

pin [pɪn] bin [bɪn] tin [tɪn] day [deɪ] kiss [kɪs] good [ɡʊd] fame [feɪm] visit [ˈvɪzɪt] thin [θɪn] rather [ˈrɑːðər] miss [mɪs] zoo [zuː] cash [kæʃ] vision [ˈvɪʒn] ham [hæm]

ʧ ʤ

match [mæʧ] bridge [brɪʤ]

m n ŋ

mum [mʌm] night [naɪt] thing [θɪŋ]

l

like [laɪk]

r [ɹ]

rise [raɪz]

j w

Consonants (19) p b t d k ɡ f

pan [pan] ‘bread’ bien [bjen] ‘well’ tan [tan] ‘so’ dos [dos] ‘two’ acto [ˈakto] ‘act’ gol [ɡol] ‘goal’ fumo [ˈfumo] ‘(I) smoke’

θ

zumo [ˈθumo] ‘juice’

s

sumo [ˈsumo] ‘(I) add’

x ʧ

jamón [xaˈmon] ‘ham chico [ˈʧiko] ‘boy’/’small’

ʝ m n

yeso [ˈʝeso] ‘plaster’ cama [ˈkama] ‘bed’ cana [ˈkana] ‘grey hair’

ɲ l ʎ

caña [ˈkaɲa] ‘cane’ lata [ˈlata] ‘tin’ llama [ˈʎama] ‘flame’

r [r̄] ɾ

carro [ˈkaro] ‘cart’ caro [ˈkaɾo] ‘expensive’

yet [jet] wet [wet]

1.6 A guide to phonetic transcription of RP and PSp The following brief notes and tips should help you to transcribe, and/or to copy transcriptions correctly, as well as to improve your pronunciation. (1) Never join symbols as in spelling. Never use capitals. (2) Mark primary stress with [ˈ] immediately before the stressed syllable, and secondary stress with [ˌ] (e.g. ˌfəʊtəˈgræfɪk). (3) Punctuation: [|] is equivalent to [,] and [||] is equivalent to [.] Never use other marks for punctuation. The apostrophe is not used to indicate contracted forms. (4) Write symbols in their correct form and clearly. Symbols that are ordinary letters must, in most cases, resemble the typed or printed form rather than certain common hand-written forms. Especially [b f g l r s v z]. In these

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

A guide to phonetic transcription of RP and PSp

39

cases Roman symbols are used as phonetic symbols to represent phonemes because they are not already in use. This explains why, although [ɹ] is the IPA symbol that represents the most common pronunciation of the initial phoneme of run in English, /r/ is used instead in phonemic transcriptions. Likewise, in addition to the five Roman letters/symbols [a e i o u] that represent the Spanish vowel phonemes, eight additional non-Roman symbols [ɑ æ ɜ ə ɪ ɔ ɒ ʊ] are also used because they are necessary to represent the vocalic qualities of RP vowels and vocalic glides. (5) Never write , , , , nor [i, u] in diphthongs, nor [r] before a consonant or after a vowel at the end of a word unless transcribing a rhotic dialect (e.g. GA) or to indicate linking with a following word-initial vowel in the same phrase (see fn 8 and Section 5.5 for further details). (6) Note that [z] represents the first sound in English , not the one in thin /θɪn/ or Spanish zoo /θoo/. However, [j] represents the first sound in yes /jes/, not the one in jet /ʤet/. (7) We can also analyse the logic of symbols. As regards consonants, the first sound in thin is made with the tip of the tongue between the teeth and the symbol [θ] is like a picture of a tongue in a mouth. [ð] retains the curved shape of a medieval scribe’s with a cross-stroke added to indicate that, unlike [d], this is a fricative sound for the production of which the tip of the tongue again protrudes between the teeth and lips. The sounds [z ʒ] are somewhat similar to each other, and the same applies to [s ʃ]. The symbol [ʒ] is a mixture of [z] and the lower part of [ɡ] which indicates the more retracted (palatal) quality of the former; whereas [ʃ] has the shape of a stretched [s]. The sound spelt consists of [t] followed by [ ʃ ], so the symbol is [ʧ]. The same applies to the voiced version [ʤ] (consisting of [d] + [z]) and sounds a bit like Spanish . The symbol [ŋ], a mixture of letters and , represents the sound spelt . As for vowels, the vowel in ten [e] is virtually the same in English and Spanish. The vowel in been [iː] is the same as Spanish [i] except that it is a long sound, as indicated by the length marks, also called triangular colons [ː]. The symbols for vowels that do not exist in Spanish are based on, or are modified versions of, the spellings of the most similar Spanish sounds. The symbol [ɪ] is a small capital ɪ. [ʌ] is a small capital A with the horizontal line omitted. [æ] is a mixture of a and e, and [ə] is e turned round so that it looks a bit like a. RP [ɑ], called script alpha, represents a much more retracted realisation than Sp [a̙]. Its rounded counterpart [ɒ], turned script ɑ, is basically the letter with a bit added, while [ɔː] is with a bit missing. [ʊ] is like u but with “wings” instead of a “foot”, and [uː], though similar in shape, sounds different from Spanish /u/ (longer and closer).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

40

Phonetics and Phonology

(8) It should also be emphasised that the distinction between “broad” and “narrow” transcription is neither a black-and-white one nor a hard-and-fast rule. We can find moderately broad and narrow transcriptions. Here syllabic consonants [C̩] will be marked both in broad and narrow transcriptions. Thus, syllabic l, for example, will be marked as /l ̩/ in broad transcriptions, but as [ɫ ̩] in narrow transcriptions. Besides syllabicity, both broad and narrow transcriptions will include, when appropriate: happy-tensing i (Section 3.2.1), linking r (Section 5.5) and weak forms (Section 5.7.1). (9) Transcriptions should always be checked in a (pronunciation) dictionary such as LPD or EPD, whose set of phonetic symbols conforms or deviates only in minor details from the one used here following Wells’ or Jones’ models. (10) As they are a tool to learn pronunciation, practise reading phonetic transcriptions aloud without referring to the orthographic transcript. Likewise, do as many transcriptions as possible and do not be discouraged if you make some mistakes. The more you practice, the fewer mistakes you will make.

Further reading The range of books and articles devoted to the pronunciation of British English is wide. Some well-known ones are Cruttenden (2014), Gimson (1984, 1994) and Kreidler (1989). The first is a revised version of the second, Gimson’s approach, and both have been two major inspirations for this book, while the last one is more focused on phonology and phonological processes. Of the early works on generative phonology, the main one is Chomsky and Halle (1968), although you may find a more accessible introduction to this approach in Brengelman (1970). Two historical books worth mentioning are: Baugh and Cable (1993) and Scragg (1974), but the latter concerns spelling rather than pronunciation. Introductory accounts of speech and phonetics (including its branches, i.e. articulatory, acoustic and auditory or perceptual) can be found in a number of textbooks such as Cruttenden (2014, Chapters 1 and 3), Ashby (2005, Chapters 1 and 2; 2011, Chapter 1), Roach (2005, Chapters 1 and 2), Ashby and Maidment (2005, Chapter 5), Ladefoged (2001, Chapter 2), Ladefoged (2003, Chapters 6 to 8), Ball and Rahilly (1999, Chapters 9 to 12), and Laver (1994, Part 1). Lodge (2009) and Katz (2013), on the other hand, offer a critical and a simplified overview of the field, respectively. For more advanced descriptions of the basic principles of (English) phonology (and phonetics), the reader is referred to Carr (2012), Clark et al. (2007), Gussenhoven and Jacobs (2005), Roca and Johnson

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

A guide to phonetic transcription of RP and PSp

41

(1999), Giegerich (1992), Katamba (1989) and Lass (1984), to mention but a few. The structure of the syllable in English is dealt with in Cruttenden (2014, Chapter 5), Roach (2005, Chapter 8), Goldsmith (1990, Chapter 3), Katamba (1989, Chapter 9) and Hogg and McCully (1987, Chapter 2). Moving on to Spanish, classic analyses of syllable structure may be found in Alcina and Blecua (1975) and RAE (1973), while other classic books that describe the sounds of PSp are, as already noted in the introduction, Quilis (1985, 1993), Alarcos LLorach (1961 [1983]), Navarro Tomás (1991[1918], 1966 [1946]). More recent approaches may be found in Martínez Celdrán and Fernández Planas (2007), Gil (2007), Hualde (2005) and Martínez Celdrán et al. (2003), whereas Colina (2009) presents Spanish phonology from a syllabic perspective. Other thorough introductions to Spanish phonetics and phonology for teachers and speakers of other languages may be found in Quilis and Fernández (1996), Núñez Méndez (2005) and Arteaga and Llorente (2009), among others. To compare the sounds of English and Spanish, some basic references also mentioned in the introduction are Stockwell and Bowen (1965), Alcaraz and Moody (1993), Finch and Ortiz Lira (1982) and Mott (2011), as well as Cabrera Abreu and Vizcaíno Ortega (2009), who propose an acoustically-based method to explain English phonetics and phonology to Spanish speakers.

Exercises 1. What is the difference between producing “noises” and “speech sounds”, i.e. between meaningless physical sounds and linguistic sounds? Think of a baby: the human being takes approximately one year before producing intelligible speech sounds, and yet s/he is not silent during that time. What makes the difference? 2. Would you justify seeing phonetics as a branch of linguistics? If so, how? Now, in a forensic phonetics investigation: how important is it that the specialist has linguistic knowledge as well – i.e. that s/he is a phonologist as well as a phonetician? Justify your answer. 3. Using the word pit as a basis, explain the difference between contrastive distribution (phonemes) and complementary distribution (allophones). 4. Think of the sequence given in Section 1.3.: pan, tan, can, ban, Dan, fan, van, ran. How would you justify that /s/ is a phoneme of English? 5. Look at the following sequences from the films ‘Pink Panther’ (the “I want to buy a hamburger” scene at http://www.youtube.com/watch?v=9q_ aXttJduk) and ‘My fair lady’ (the ‘Accent training’ scene at http://www. youtube.com/watch?v=l8CID-bXgo4&feature=related). Comment on Inspector

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

42

Phonetics and Phonology

Clouseau and Eliza Doolittle’s training sessions aiming at good pronunciation: are these sessions phonetic or phonological in nature? 6. Given the following two utterances, in English and in Spanish, identify those phonemes that are likely to sound foreign when pronounced by a nonnative speaker. Explain in your own words why they would sound foreign. (1) We have some ham sandwiches ready for tea in the garden (2) El perro está en la jaula y no quiere comer. 7. Give the syllabic pattern of the following words: three, tax, table, awe, checking. 8. Provide one monosyllabic word for each of the RP English phonemes below, using minimal pairs where possible: /æ ɑː ɜː ɪ ɔː ʊ əʊ eə t g v θ ʃ ʤ h r j/ 9. Give the phonetic symbol as required: initial consonant medial consonant final consonant vowel

1. car 5. cushion 9. sing 13. war

2. school 6. button 10. axe 14. blood

3. yes 7. branches 11. faith 15. earn

4. this 8. over 12. bottle 16. kiss

10. Initial consonant clusters. The following consonant clusters do not occur in Spanish. Pronounce the following words, paying special attention to the initial consonant clusters. Transcribe them. /θr/: threat three thrill through thrive throat throw /ʃr/: shred shriek shrimp shrug shrine 11. /s/-initial clusters are particularly difficult for Spanish speakers. Pronounce the following words trying not to insert an extra /e/ sound before the initial cluster. Transcribe them. /sk/: scab sketch skin score school skirt skull skate scare scout square squeeze /sl/: slam slender sleep slob slew slurp slave slide slow slogan /sm/: smart smell smith smog small smile smoke /sn/: snap sniff sneeze snob snail snare snow /sf/: sphere /sp/: spam spa special spend speak spark spin spirit speed speech sport spoon Spain spare spice spoil spoke /st/: stack stand star steady stem steam stick still steal stock stop storm stool student stir study stay stile stone /skr/: scrap script screw scream screen

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

A guide to phonetic transcription of RP and PSp

43

/spl/: splash split splendid /spr/: sprang spread spring spray /str/: strap stress street strict strong strange strike 12. Final consonant clusters are particularly difficult for Spanish speakers, especially those ending in a plosive. Pronounce the following words carefully, paying special attention to the endings. Then transcribe them. /ft/: draft left gift soft /kt/: act affect depict deduct /lt/: belt built fault /ld/: build field old cold gold told child wild world /nt/: can’t sent tent mint want hunt saint pint point /nd/: hand send friend wind pond fund find found sound pound 13. Render the following passage into ordinary spelling: / fəˈnetɪks ɪz ə brɑːntʃ əv lɪŋˈɡwɪstɪks ðət kəmˈpraɪzɪz ðə ˈstʌdi əv ðə ˈsaʊndz əv ˈhjuːmən spiːtʃ / ɔːr ɪn ðə keɪs əv saɪn ˈlæŋɡwɪdʒɪz ði ɪˈkwɪvələnt ˈæspekts əv saɪn // ɪt s kənˈsɜːnd wɪð ðə ˈfɪzɪkl ̩ ˈprɒpətɪz əv spiːtʃ ˈsaʊndz ɔː saɪnz / fəʊnz / ðeə ˌfɪzɪəˈlɒdʒɪkl ̩ prəˈdʌkʃn̩ / əˈkuːstɪk ˈprɒpətɪz / ˈɔːdɪtəri pəˈsepʃn̩ / ənd ˈnjʊəˌrəʊˌfɪzɪəˈlɒdʒɪkl ̩ ˈsteɪtəs // fɒˈnɒlədʒi / ɒn ði ˈʌðə hænd / ɪz kən ˈsɜːnd wɪð ði ˈæbstrækt / ɡrəˈmætɪkl ̩ ˌkærəktəraɪˈzeɪʃn̩ əv ˈsɪstəmz əv ˈsaʊndz ɔː saɪnz //

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:02 PM

Chapter 2

2 The Production and Classification of Speech Sounds 2.1 Introduction This chapter describes the way human speech is produced (Section 2.2), what articulatory parameters are generally used to classify vowels (Section 2.3.1) and consonants (Section 2.3.2), as well as what the acoustic features of (RP, PSp) speech sounds are (Section 2.4).

2.2 The organs of speech All the organs involved in the production of speech sounds can be arranged into three groups, or systems: the respiratory system, where the initial breathing process is initiated; the phonatory system, where vibration, or phonation, takes place, and the articulatory system, where resonance is modified in the VT, as illustrated in Figure 11.

Figure 11: Systems involved in the production of speech

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

The organs of speech

45

Three main airstream mechanisms or ways of initiating airflow to produce sounds are recognised depending on the source of the airstream: pulmonic if the source is in the respiratory system (the lungs), glottalic when the airflow proceeds from the phonatory system and velaric if the airstream is generated in the articulatory system (Pike 1943; Abercrombie 1975; Clark et al. 2007). A further distinction involves the direction of the airstream: if the air flows outward, then egressive sounds are produced, while ingressive sounds are those in which the airstream flows inward through the mouth or nose. Speech sounds in English and Spanish are produced with an eggressive (or outgoing) pulmonic airstream (or column of air), which moves upwards from the lungs (breathing) through the larynx (phonation) and outwards through the vocal tract cavity (resonance) consisting of the pharyngeal, oral and nasal cavities, where a series of articulators come into play. None of these organs conforming the three systems involved in speech production have speech as their main original function (e.g. the lungs are for breathing, the vocal cords are for preventing choking, the tongue is for eating and tasting, the nose for breathing and smelling, and so on), but they have been adapted to produce communicative sounds, as will be explained in what follows.

2.2.1 The respiratory system and pulmonic sounds The main organs in the respiratory system are the lungs, which are connected with the exterior by means of the bronchial tubes and the trachea (or windpipe). The lungs are contained within the thoracic cavity, protected by the rib cage and separated from the abdominal cavity by the diaphragm. The breathing cycle (Figure 12) involves the expansion and contraction of the lungs in a controlled manner, a process which normally takes four seconds. The rib cage expands by lowering the diaphragm, which results in air flowing into the lungs (inspiration or inhalation). After filling with air, the lungs collapse under their own heavy weight, so the diaphragm muscle is raised and the chest contracts so that an outgoing flow of air (expiration or exhalation) is started.

AI 2.1 Ingressive and egressive airstream

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

46

The Production and Classification of Speech Sounds

Figure 12: The breathing cycle

In many languages, such as English or Spanish, as already noted, all the speech sounds are articulated during the exhalation process with eggressive pulmonic airstream. This means our utterances are partly shaped by the physiological capacities imposed by our lungs and by the muscles that control their actions. When we speak we are forced to make pauses in order to refill our lungs with air and this will to some extent determine the division of speech into intonational phrases (see the discussion of tonality in Section 6.3). Pulmonic ingressive sounds are also possible. They are produced when the airflow is going into the lungs, which alters the voice quality considerably. But it is unclear whether ingressive pulmonic sounds occur as normal speech sounds. Fuller (1990) claims to have recorded one speaker of Tsou, an Austronesian language of Taiwan, using pulmonic ingressive fricatives in word-initial position, but Ladefoged and Zeitoun (1993) were unable to attest this with other speakers from the same village. The only cases of pulmonic ingressives as normal speech sounds are apparently those used by women with other women in certain situations in Tohono O’odham (Papago) (Hill and Zepeda 1999). It seems that pulmonic

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

The organs of speech

47

ingressives (inhaled speech) are mostly employed paralinguistically,13 as in the case of Japanese ingressive [s] (produced when the speaker is upset), the Scandinavian languages (usually with feedback words (yes, no) or cries of pain or sobbing), the English ingressive interjection heuh! (used to express surprise or empathy when someone is hurt), in Portuguese (also in interjections) or in Brazilian falar para dentro (‘talking to the inside’) (produced when speakers talk to themselves when they are alone or manifesting discomfort). Also, when we snore we produce a sound with an ingressive pulmonic airstream.

2.2.2 The phonatory system, phonation modes and glottalic sounds The phonatory system includes the laryngeal structures through which phonation is achieved, regulating the air flow to create both voiced and voiceless segments (in addition to other phonation types), and it is the source of air pressure used to produce glottalic sounds. The larynx, colloquially known as the voicebox or Adam’s Apple, is a casing ring situated at the top of the trachea that consists of nine separate cartilages and is bigger in males than in females (Figures 13 and 14). Through the use of certain laryngeal muscles, it can be moved slightly upwards or downwards, producing different voice quality effects and aiding in the process of bringing up air from the lungs and through the trachea. Within the larynx, running from the arytenoids forward to the interior of the front of the thyroid cartilage, are the vocal folds, or more commonly vocal cords, two whitish bands of ligament that are typically about 17 to 22 mm long in males and about 16 mm in females (Clark et al. 2007: 178; Ball and Rahilly 1999: 8–11). The action of the vocal cords is controlled by forward, backward and sideto-side movements of the two arytenoid cartilages. Backward and forward movements of the arytenoid cartilages adjust the tension of the vocal cords. The more tense the vocal folds are, the higher the perceived pitch of speech sounds. In contrast, side-to-side movements of the arytenoids achieved by using the posterior cricoarytenoid muscles, either separate (or abduct) or bring

13 Paralanguage refers to the conscious or unconscious use of non-verbal elements (gestures, giggling and the like) to modify meaning, convey emotion or signal an attitude or a social role. The term paralinguistics is restricted to vocally produced sounds or variations in tone of voice (with breathy or creaky voice or by adopting secondary articulations such as nasalizations or labializations) which produce the same effect and seem to be less systematic than prosodic features (intonation and stress) (Crystal 2008: 349).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

48

The Production and Classification of Speech Sounds

Figure 13: Larynx in cross section

Figure 14: Larynx viewed from above

together (or adduct) the vocal folds. When the vocal folds are abducted, or held wide apart, as shown in Figure 15 (a), a clear passage for the airstream is allowed so that no vibration occurs. This is what happens in normal breathing

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

The organs of speech

49

or when voiceless consonants are produced. If the vocal folds are adducted, or brought together (or nearly so) along most of their length, as in Figure 15 (b) below, they vibrate when the airstream passes through them producing voice, as it occurs in the production of vowel sounds and voiced consonants. In addition, the vocal folds may be set together leaving a small opening between them.

Figure 15: Human larynx and vocal folds

Normally, the vocal cords come together rapidly but part more slowly, so that the opening phase is comparatively longer than the closing stage. The space that is left between the vocal folds is called the glottis, or the glottal slit, which may be opened or closed to varying degrees. Considering the different degrees of aperture and tension of the glottis, seven main phonation modes can be recognised, as represented in Figure 16 below: (a) voice or modal voice, (b) unvoiced or pulmonic, (c) glottal stop, (d) creak, (e) creaky voice, (f) whisper, and (g) breathy voice. The first three apply to the production of certain individual sounds of the language, whereas the last four refer to the whole chain of connected speech, regardless of the various voiced, voiceless or glottal sounds in it. But in all seven phonation types, eggresive pulmonic airflow passes through the glottis within the larynx, so that a series of modifications take place involving the vocal folds, the arytenoids and other laryngeal muscles. These seven phonation types are explained in what follows, but they can be best observed with a laryngoscope, which gives a stationary mirrored image of the glottis, or through stroboscopic techniques, which allow to obtain a moving record and high speed films of the vocal cords in action.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

50

The Production and Classification of Speech Sounds

Figure 16: Phonation modes

The mechanism of voicing (voice) reflects the so-called Bernoulli principle, according to which a moving stream of gas or liquid tends to pull objects from the sides of the stream to the middle. The faster the stream goes, the stronger the pull. In voicing, the vocal folds are held fairly close together, as shown in Figures 15 (b) and 16 (a) above. When the pulmonic airstream passes between them, the Bernoulli effect, together with the elastic tension of the folds, pulls them together. As soon as the vocal folds are together, the Bernoulli effect ceases and the force of the airstream from below pushes them apart again, but as soon as they are apart, they are pulled together again as a result of Bernoulli effect, and so the process continues. Voiced phonation, then, involves expelling short puffs of air very rapidly by the repeated vibration of the vocal folds. The rate of these vibrations controls the fundamental frequency of a sound, which is, on average, 120–130 times per second in an adult male speaker, and about 220–230 times per second for an adult female. This determines what we perceive as pitch, whereby sounds are recognised as being high or low: the faster the vibration of the vocal folds, the higher the pitch of the sound. Slow vibration, resulting in a deeper pitch, may result from longer and larger vocal folds as in the bigger larynxes of males. Voicelessness (adjective unvoiced, but also pulmonic) is a specific adjustment of the glottis and not just the absence of voicing. It refers to the abduction of the vocal folds that results in the opening of the glottis, as represented in Figures 15 (a) and 16 (b) above. The vocal cords (and the arytenoids) are open at between 60% and 95% of its maximal opening, and the pulmonic eggressive airstream flows relatively freely through the larynx. This kind of airflow is characterised by nil phonation. All languages have both voiceless and voiced sounds contrasting in their phonological systems. Interestingly enough, in most European languages, like English and Spanish, voiced sounds are in general three times more common than voiceless ones; but other languages may have ratios that are more balanced (Dutch), or even voiceless sounds occurring more often than voiced ones (Korean). In English and Spanish, sonorants and vowels are rarely voiceless, whereas obstruents are commonly found voiceless, although they can also occur voiced.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

The organs of speech

51

Somewhat resembling a weak cough or a cork pulled out of a champagne bottle, the glottal stop [ʔ], illustrated in Figure 16 (c) above, is produced by bringing together the vocal folds (and the arytenoids) blocking the airstream coming from the lungs behind them for a moment, followed by the sudden release of this pressurised air. In some languages, glottal stops are actual phonemes, as in Arabic and Persian, while in English they can occur both segmentally (e.g. as a realisation of final [p], [t] and [k], as in clap, what and cock) and prosodically (e.g. as a hiatus blocker or pause marker in co-operate) (see § 5.2.8.2 for further details).

AI 2.2 Glottal stop In addition, there exist other glottalic sounds that are produced with an airstream coming from or to the larynx, by closing the vocal folds tightly shut so that no air can pass through the glottis. Glottalic egressive sounds (produced with an outgoing airstream and a closed glottis as well as with a supralaryngeal closure gesture) are called ejectives and they are especially common in the native languages of North America (Pacific Northwest) but are very infrequent in Europe (except in the Caucusus region, at the border of Europe and Asia). Less common than ejectives, glottalic ingressive sounds (produced with an ingoing airstream) are called implosives and they are especially common in Africa and Central America (Mayan languages). Creak, also termed glottal fry or vocal fry because of the sputtering effect it produces, consists of pulses of air passing through the glottis with the arytenoids tightly closed, allowing only the front portion of the vocal folds to vibrate and producing a succession of glottal stops, as shown in Figure 16 (d) above. Creak has low sub-glottal pressure and low volume velocity air flow; and the frequency of vocal fold vibration can be in the region of 30–50 pulses per second. Creaky voice, represented in Figure 16 (e), combines creak with voice. It is often used by English speakers paralinguistically (replacing modal voice) either to suggest boredom, authority, avoid disturbing people, or to keep a conversation private. Some people use this voice idiosyncratically as a sign of affectation.

AI 2.3 Creaky voice Whisper, also known as library voice, requires the glottis to be closed by about 25% and the vocal folds to be closer together than for voicelessness, especially the anterior section of the folds, whereas the triangular-shaped opening

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

52

The Production and Classification of Speech Sounds

takes place at the back, so that a considerable amount of air escapes at the arytenoids, as illustrated in Figure 16 (f). Air flow is strongly turbulent, which produces the characteristic hushing quality of whisper. This phonation type is used contrastively in some languages, but we are more used to thinking of whisper as an extra-linguistic device to disguise the voice or, at least, to reduce its volume. Breathy voice, in Figure 16 (g), is a combination of whisper and voice. Although the vocal cords are open, the expulsion of air is so strong that they are made to vibrate. This is the voice associated with “sexy” voices, and is also known as “bedroom voice”, sometimes used by singers as a special effect. Finally, we should mention falsetto voice, which is produced when the thyroarytenoid muscle contracts to hold the vocal folds very tightly, allowing vibration at the edges. The glottis is kept slightly open and sub-glottal pressure is relatively low. The resultant phonation is characterised by very high frequency vocal fold vibration (between 275 and 634 pulses per second for an adult male). Falsetto is not used linguistically in any known language, but has a variety of extralinguistic functions dependant on the culture concerned (e.g. greeting in Tzeltal, Mexico). Falsetto voice is pertinent to males, as women’s voices are generally higher pitch anyway, and is used in singing more often than in speaking.

2.2.3 The articulatory system and velaric sounds The articulatory system, also known as vocal tract (VT), consists of various elements distributed in three supralaryngeal, or supraglottal, cavities that are illustrated in Figure 17: the pharynx (throat in everyday language), the nasal cavity (nose) and the oral cavity (mouth), which act as resonators and alter the sound produced by the vibration at the vocal folds, by providing the necessary amplification or diminishing it. Sounds, particularly the vowels, can also be modified in these cavities by the alterations in shape which they can adopt, and also, particularly the consonants, by means of the various articulators within: the soft palate or velum, the hard palate, the alveolar ridge, the tongue, the teeth and the lips. A description of each of these three articulatory cavities is provided in turn. In addition, at the end of this section, we shall also see that the articulatory system, or to be more precise the tongue, is the source of air pressure that is necessary to produce velaric sounds. Forming the rostral boundary of the larynx, the epiglottis is a small movable muscle whose function is to prevent food from going down the trachea into the lungs and so divert it to the oesophagus down to the stomach. The pharynx is a tube about 7 to 8 cm long, which runs from the top of the larynx up to the

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

The organs of speech

53

back of the nasal and oral cavities. The lower area of the pharynx, nearer the oral cavity, is called oropharynx, and the upper part, nearer the nasal cavity, is the nasopharynx. The section of the pharynx which is immediately above the larynx is called laryngopharynx, and the epilarynx is the area of contact between them. The size and shape of the pharynx can be altered in a number of ways. The overall volume of the pharynx may be reduced, by raising the larynx, or it may be reduced besides adding an obstruction to the airflow, if the root of the tongue and the epiglottis are retracted into the oropharynx, or even its back wall can be contracted, by drawing the side walls to each other. These different articulatory gestures result in corresponding alterations of the quality of the voice. Access to the nasal cavity, which can be touched by the back of the tongue, is triggered at the velopharyngeal port by lowering the soft palate, or velum, so that the outgoing body of air is released through the nares, or nostrils (although it may also simultaneously flow through the oral cavity), and nasal resonance is provided. This is what happens in the articulation of nasal and nasalised sounds. But in normal breathing and to produce oral speech sounds the velum is raised, so that the airstream is liberated through the mouth without entering the nasal cavity.

Figure 17: The articulatory system

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

54

The Production and Classification of Speech Sounds

Let us focus on the oral cavity, the most versatile of the three supralaryngeal cavities. The mouth may be closed or opened by raising or lowering the lower jaw, or mandible. The upper and lower lips are flexible and can adopt a variety of positions. They can be in a neutral shape, or they can be open (or held apart), closed (or brought together), or rounded in different degrees. So we can say, for instance, that the lips are either closely rounded, or tightly rounded (vs. slightly rounded or loosely rounded), or spread apart (either loosely or tightly). They can also come into contact with the teeth, which are fixed in position and act as obstacles to the airstream. The tongue, on the other hand, is the most flexible of the articulators within the supralaryngeal system. It can adopt many different shapes and can also come into contact with many other articulators. The tip (adjective apical) and blade (adjective laminal) can either approximate or touch the upper teeth, and the alveolar ridge, the section between the upper teeth and the hard palate; but they can also bend upwards and backwards so that their underside can touch the roof of the mouth, or hard palate,14 which can also be touched by the front of the tongue. The back of the tongue can be raised against the velum and the uvula, whereas its root (or base) can be retracted into the pharynx. The area where the front and back of the tongue meet is known as centre (adjective central). Likewise, the front, centre and root of the tongue are sometimes collectively known as the body of the tongue, while the edges of the tongue are called rims. We shall see that sound descriptions necessarily refer to (1) the height of the tonge, that is, whether it is raised or touches the teeth, alveolar ridge, and so on, because different places of articulation produce different sounds, and (2) the position of the tongue in the mouth, that is, whether it is advanced or retracted, which affects the size of the oropharyngeal cavity, and consequently, influences the quality of the sounds produced (especially vowels). Before closing this section, let us consider velaric sounds. These are made by the body of the tongue trapping a volume of air between two closures in the mouth, one at the velum (the back of the tongue is placed against the soft palate), and one further forward (the tip, blade and rims of the tongue are placed against the teeth and the alveolar ridge). Velaric egressive sounds (produced with an outgoing airstream) are physically impossible because it is not possible to compress the portion of the oral tract between the velar closure and the anterior closure. Velaric ingressive sounds (produced with an ingressive airstream) are called clicks and tend to be used paralinguistically (mainly as 14 Palatography and electropalatography, studying the kind and extent of the area of contact between the tongue and the roof of the mouth, provide a practical way of recording tongue movements and illustrating the articulation of speech sounds.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

55

interjections). In English and Spanish, the kissing sound that people make is a bilabial click ([ʘ]), whereas the alveolar click [!] is used to express disapproval or annoyance and the velar click [ǁ] is the sound produced to encourage horses. The only languages that use clicks as regular speech sounds are found in Southern Africa, to be more precise the Khoi and San languages, as well as some of the Southern Bantu languages.

2.3 Articulatory features and classification of phonemes When the egressive pulmonic air passes through the phonatory system and reaches the articulatory system of the oral and nasal cavities, it is modified by certain organs that move against others and may be released in different ways, depending on the degree of aperture of the mouth. In this section we shall see that vowels, vowel glides and consonant sounds are produced differently and therefore need different parameters of classification.

2.3.1 Vowels and vowel glides A complete characterisation of vowels, or vocalic sounds, and vowel glides involves three types of features: (1) functional (or phonological), (2) acoustic/ auditory and (3) articulatory. Functionally, vowels are syllabic, that is, they are the nucleus of the syllable, thereby getting intonational prominence (unlike semivowels or semiconsonants15 and approximants (see Sections 2.3.2.3 and 4.2.4)), which tend to be marginal in the syllable. From an acoustic point of view, vowels and vowel glides are characterised by homogenous and regular formant structure patterns, as will be further discussed in Sections 2.4.1 and 2.4.2. Articulatorily speaking, vowels are characterised by having no obstruction in the VT: the air-stream comes through the mouth (or through the mouth and nose), centrally over the tongue and meets a stricture of open approximation, in other words, there is a considerable space between the articulators in their production. Seven other articulatory features determining vowel quality are 15 The terms semi-consonant and semi-vowel may be used interchangeably although they bring different ideas to the fore. The label semi-consonant highlights the consonantal quality of segments that function as a syllabic margin (e.g. English [j] and [w] in [j + é], yet /jet/, [j + ɪ] year /jɪə/, [w + é] wet /wet/) but are not the nucleus or peak (i.e. the most prominent or sonorous part) of the syllable. In contrast, the label semi-vowel reinforces the idea that the segment has the phonetic (articulatory, auditory and acoustic) characteristics of a vowel, but the phonological behaviour of a consonant (it occurs in syllable margins) (Crystal 2008: 431).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

56

The Production and Classification of Speech Sounds

observed in relation to the action of the vocal folds, the soft palate, the tongue and the lips, as well as the muscular effort employed in their articulation: (1) The action of the vocal cords during phonation: generally all vowels and vowel glides are voiced (i.e. produced with vibration of the vocal folds), but they may be devoiced, especially when occurring next to a voiceless plosive (as in the second vowel of carpeting or multiple) (for more details on the devoicing of RP vowels, see § 5.2.2); (2) The action of the velum or soft palate: raised, in oral articulations, or lowered, in nasal(ised) articulations;16 generally all vowels in RP and PSp are oral (i.e. the air escapes through the mouth), but they can be nasalised, especially if followed by nasal consonants (like [e] in ten [tẽn] or [a] in PSp pan [pãn] ‘bread’) (for further details on the nasalisation of RP vowels see § 5.2.4); (3) Tongue height, which refers to how close the tongue is to the roof of the mouth, and consequently determines the degree of openness of the mouth and of the vowels according to four values: close (high), half-close (high-mid), half-open (low-mid) and open (low). This parameter is further discussed in Section 2.3.1.1. (4) Tongue backness, which refers to the part of the tongue that is highest in the articulation of a vowel (tip, blade, front or back, see Fig. 16 above) rendering three values of vocalic description: front, central and back. These values are further explained in Section 2.3.1.1. (5) Lip shape, basically involving three positions (slightly/tightly) rounded, spread and neutral, as will be explained in Section 2.3.1.2. (6) Duration, or the length of the vowel, and energy of articulation, that is, the muscular effort required to articulate a vowel, more details of which are given in Section 2.3.1.4. (7) Whether vowel quality is relatively sustained (i.e. the tongue remains in a more or less steady position), or whether there is a transition or glide from one vocalic element to another (or others) within the same syllable, as will be noted in Section 2.3.1.5. In what follows, further details are given about the articulatory features and classification of vowels and vowel glides, as well as about their relation to the system of cardinal vowels (Section 2.3.1.3). Chapter 3 (Sections 3.2 and 3.3) offers 16 The terms ending –ised (adj.)/–isation (n.) generally refer to a secondary articulation (see § 2.3.2.5.). That is, any articulation which accompanies another (primary) articulation and which normally involves a less radical constriction than the primary one (e.g. nasalised-nasalisation, palatalised-palatalisation, etc.).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

57

a detailed description of the vowels and vowel glides of RP in comparison to those of PSp, while Chapter 5 summarises their main realisational or allophonic variants.

AI 2.4 Vowels and glides of English RP 2.3.1.1 Tongue shape There are two parameters involved in vowel articulation that concern the tongue: tongue height and tongue backness. The position of the highest point is used to determine vowel height and backness. Tongue height indicates how close the tongue is to the roof of the mouth. If the upper tongue surface is close to the roof of the mouth (like /iː/ in fleece and /uː/ in goose), then the sounds are called close vowels, or high vowels. By contrast, when vowels are made with an open mouth cavity, with the tongue far away from the roof of the mouth (like /æ/ in trap and /ɑː/ in palm), then they are termed open vowels, or low vowels. There are two further intermediate values between these two: half-close (highmid) and half-open (low-mid), which represent a vowel height between close and half-open in the former case, and between open and half-close in the latter (see also Section 2.3.1.3 on cardinal vowels). In RP there are four close/high vowel phonemes /iː ɪ ʊ uː/, three open/low vowels /æ ɑː ɒ/ and five mid vowels /e ʌ ɜː ə ɔː/; while PSp has two close/high vowel phonemes /i u/, one open/low vowel /a/, and two mid vowels /e o/. The degree of openness or closeness of vowels may be further specified by means of two diacritics: [˔] (bit [bɪ ̝t]) and [˕] (sit [sɪ ̝t]), which indicate respectively raised (closer) or lowered (more open) realisations of vowels within these four values. Tongue backness, in turn, identifies which part of the tongue is highest in the articulation of the vowel sound: if the front of the tongue is highest, we speak of front vowels (like /iː/ in fleece); if the back of the tongue is the highest part, we have what are called back vowels (like /ɔː/ in cord or /uː/ in clue). Central vowels are articulated with the tongue in a neutral position, neither pushed forward nor pulled back, but it may be raised to the degrees mentioned above (like /ə/ in the second syllable of venom, which represents a central vowel between half-open and half-close). In RP there are three central vowels /ʌ ə ɜː/, four front vowel phonemes /iː ı e æ/, and five back vowel phonemes /ɑː ɔː ɒ ʊ uː/, whereas PSp has two front vowels /i e/, one central vowel /a/ and two back vowels /o u/. The degree of frontness or backness of vowels may be further specified by means of the diacritics [+] [–], which express more retracted or more advanced vocalic realisations within these four values. Both are usually placed above the vowel symbol, but they may also follow it, as will be illustrated in Chapter 5.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

58

The Production and Classification of Speech Sounds

Testing RP close and open vowels To test close and open vowels, say the English vowel /ɑː/ as in palm. Put your finger in your mouth. Now say the vowel /iː/ as in fleece. Feel inside your mouth again. Look in a mirror and see how the front of the tongue lowers from being close to the roof of the mouth for /iː/ to being far away for /ɑː/. Now say these English vowels: /iː/, /ɜː/ and /æ/. Can you feel the tongue moving down? Then say them in the reversed order and feel the tongue moving up. Testing RP front and back vowels To test front and back vowels, take another set of English vowels: /ɑː/ and /ɔː/ and /uː/. Notice how it is the back of the tongue that raises for /ɔː/ and /uː/, whereas for /ɑː/ the tongue is fairly flat.

2.3.1.2 Lip shape The second parameter used to describe different vowel qualities is the shape of the lips. We will consider mainly three possibilities: (1) tightly or slightly rounded, or pursed: the corners of the lips are brought towards each other and the lips pushed forwards ([u]); (2) tightly or slightly spread: the corners of the lips are moved away from each other, as for a smile ([i]); and (3) neutral: the lips are not noticeably rounded or spread – as in the noise most English people make when they hesitate, spelt er. The main effect of lip-rounding is the enlargement of the mouth cavity and the decrease in size of the opening of the mouth, both of which deepen the pitch and increase the resonance of the front oral cavity. Lip shape affects vowel quality significantly. A typical pattern is found in most languages of the world, whereby front and open vowels have spread to neutral position, whereas back vowels have rounded lips (although reverse positions are also possible, as in the French vowel in neuf, for example). In RP all front and central vowels are unrounded, while all back vowels (except /ɑː/) are rounded, and the same applies to PSp. This seems to be the general tendency according to which every language has at least some unrounded front vowels and some rounded back vowels. Lip rounding makes back vowels sound more different from front vowels and have greater perceptual contrasts. In addition, it should be noted that labialised variants of consonants occur (annotated with a superscript [ʷ]) in the vicinity of a rounded vowel, as in the /p/ and /t/ of put [pʷʊtʷ]. Further details on the lip positions of RP and PSp vowels, as well as on the phenomenon of labialisation, are offered in Chapters 3 and 5 respectively.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

59

2.3.1.3 The cardinal vowels and the Cardinal Vowel Scale We have established the closest and most front vowel [i]; the closest and most back vowel [u]; the most open front vowel [a] and the most open and most back vowel [ɒ]. These degrees of aperture plus the front-back distinction define 8 reference points based on a combination of articulatory and auditory criteria, which represent eight different vocalic qualities represented in Figures 18 (a) and 18 (b).

Figure 18: Highest tongue positions for (a) cardinal front and (b) cardinal back vowels

The cardinal vowels (CV) (termed after the cardinal points of the compass) indicate the upper and lower vowel limits beyond which the tongue cannot rise, in relation to the roof of the mouth, without friction or tongue depression, as shown in Figure 19 below. The vowels at the upper vowel limit are the front vowel [i] and the back vowel [u]; the vowels at the lower vowel limit are the front vowel [a] and the back vowel [ɒ].

Figure 19: Relative highest and lowest points of the tongue for cardinal vowels

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

60

The Production and Classification of Speech Sounds

Taking the aforementioned into account, in 1917 the English phonetician Daniel Jones devised the Cardinal Vowel Scale (CVS), also called Cardinal Vowel Quadrilateral, or Cardinal Vowel Trapezium, which is presented in Figure 20 below (see also the vowel matrix on the IPA, reproduced as Table 3 in Section 1.4.1).

Figure 20: The Cardinal Vowel Scale

The primary cardinal vowels, CV 1 to CV 8 in Figure 20 above, should be distinguished from corresponding secondary cardinal vowels, which are derived from corresponding primary cardinal vowels by reversing the lip-rounding, from CV 9 to CV 18 which are represented in grey in Figure 21 below. Thus, primary CVs 1 to 5 are unrounded, and 6 to 8 are rounded; whereas secondary CVs 9 to 13 are rounded, and 14 to 16 unrounded.

Figure 21: Primary cardinal vowels vs. secondary cardinal vowels

Table 5 below revises the identification and description of the primary cardinal vowels in terms of tongue height and lip shape. There is no agreed abbreviation to describe vowels, but our habit is to speak of OBR labels (openness, backness

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

61

and lip rounding). Notice also that cardinal vowels do not refer to the vowels of a particular language, but rather are a set of reference vowels used by phoneticians in describing the sounds of languages in general. In Chapter 3 the vowels of RP will be allocated in the relevant places of the CVS. Table 5: Primary Cardinal Vowels CV

IPA

Identification

Description

1

[i]

close front unrounded

At the top left corner of the CVS, CV 1 stands for an articulation in which the lips are unrounded, and the front of the tongue is located as high as possible.

2

[e]

close-mid front unrounded

It stands for a more open articulation than CV 1, in which the lips are also unrounded, but the front of the tongue is lowered somewhat from the CV 1 position.

3

[ɛ]

open-mid front unrounded

It stands for a more open articulation than CV 2, keeping the same lip shape and degree of frontness, but with the tongue lowered to the half-open position.

4

[a]

open front unrounded

It stands for an even more open articulation than CV 3, in which the lips remain unrounded, but the tongue is moved as far to the front of the vowel space as possible while remaining as low as possible.

5

[ɑ]

open back unrounded

At the bottom right-hand corner of the CVS, it stands for an articulation in which the lips are again unrounded and the body of the tongue is again as low as possible, but as far back as possible in the vowel space, without causing friction.

6

[ɔ]

open-mid back rounded

Immediately above CV 5, CV 6 stands for a closer and rounded articulation. The lips are rounded and the tongue is raised as far back as possible to the half-open position in the vowel space, without causing friction.

7

[o]

close-mid back rounded

Immediately above CV 6, CV 7 retains the same degree of backness and a higher degree of lip-rounding, but the tongue height is raised to the half-close position.

8

[u]

close back rounded vowel

At the top right-hand corner of the CVS, CV 8 stands for an articulation in which the lips are tightly rounded and the tongue is as high and as far back as possible without causing friction.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

62

The Production and Classification of Speech Sounds

2.3.1.4 Duration and energy of articulation Duration means the time each sound takes to be pronounced, which is only of linguistic significance if the relative duration of sounds is considered. The pace of delivery in the production of speech sounds is auditorily perceived as length, involving in the case of vowels the short and long distinction (vocalic quantity). In PSp this difference does not entail a phonemic contrast in the vocalic system, but in other languages the duration of the production of a vowel has a phonemic contrast, which is often combined with vowel quality, and this is the case of RP (see § 3.2 for further details). In RP there are five long vowels /ɑː ɜː iː ɔː uː/ and seven short vowels /æ ʌ e ɪ ə ɒ ʊ/. But the relative duration of a long phoneme may be lengthened or reduced, depending on the phonetic context in which it occurs. In the first case we speak of (extra) lengthening and it is indicated with double length marks [ːː], as in the realisation of [iː] in tea [tiːː], especially when the word is emphatic; whereas cases of vowel length reduction are referred to under the umbrella term clipping, which is marked with only a single length mark or triangular colon [ˑ], as in the realisation of [iː] in leap [liˑp]). For further details on vowel allophones involving differences in length, the reader is referred to Sections 2.3.2.4 and 5.2.1. Now turning to the amount of muscular tension required to produce vowels, if they are articulated in extreme positions, they are more tense (like /iː/ in tea or /uː/ in blue) than those articulated nearer the centre of the mouth which are lax (like /ə/ in the second syllable of venom). In RP the five long vowels are tense /ɑː ɜː iː ɔː uː/ and the remaining short vowels are lax /æ e ɪ ə ʌ ɒ ʊ/, while in Spanish all vowels are tense (Monroy Casas 1980, 1981, 2012) (see § 3.2 for further details). SSLE should know that in English both tense and lax vowels can occur in closed syllables, but (apart from unstressed vowels) only tense vowels can occur in open syllables (Ladefoged 2001).

2.3.1.5 Steadiness of articulatory gesture A final classification of vowel sounds involves the steadiness of the articulatory gesture adopted in vowel production. If the positions of the tongue and lips are held steady during production of a vowel sound, the resulting sound is known as a steady-state vowel, pure vowel, or monophthong. As already seen in Table 4, in RP there are twelve pure vowels /æ e ɪ ə ʌ ɒ ʊ ɑː ɜː iː ɔː uː/, which in Chapter 3 (Section 3.2) will be further described and compared with the five vowels of PSp /a e i o u/. If there is a clear change, or glide, in the tongue or lip shape we speak of diphthongs or triphthongs, in which the glide is carried out in one single

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

63

movement from one vocalic position to another within the syllable. As already displayed in Table 4, in RP there exist eight diphthongs /aɪ eɪ ɔɪ aʊ əʊ eə ɪə ʊə/ and five triphthongs /eɪə aɪə ɔɪə əʊə aʊə/, which in Chapter 3 (Section 3.3) receive independent treatment and are compared with the fourteen diphthongs /ai ̯ ei ̯ oi ̯ au̯ eu̯ ou̯ wa we wo wi ja je jo ju/ and four triphthongs /wei ̯ wai ̯ jei ̯ jai ̯/ of Psp.

2.3.2 Consonants While vowels are produced with no obstruction to the airflow as it passes through the VT, for the articulation of consonants this obstruction does exist (Hooke and Rowell 1982). Consonants are identified according to three-term VPM labels standing for Voice, Place and Manner of articulation, which together give the consonant its distinctive quality. The voiced/voiceless opposition, as well as the fortis/lenis contrast are related to phonation, and are discussed in Section 2.3.2.1. Place of articulation and manner of articulation are explained in Sections 2.3.2.2 and Section 2.3.2.3, respectively, the latter including the subservient feature of orality, which is described separately in Section 2.3.2.4. Thus, in order to describe and classify consonant sounds according to production, we have to speak of these four parameters, in addition to that of secondary articulation, which is discussed in Section 2.3.2.5. From an acoustic point of view, consonants may be broadly characterised as having non-homogenous and irregular formant structure patterns, as will be further discussed in Section 2.4.3. A detailed description of RP consonants as compared to those of PSp is offered in Chapter 4, while Chapter 5 summarises their main realisational or allophonic variants.

2.3.2.1 Voicing and energy of articulation Voice is a fundamental term used in the phonetic classification of speech sounds, referring to the auditory result of the vibration of the vocal cords. Sounds produced while the vocal cords are vibrating are voiced sounds; those produced with no such vibration are voiceless or unvoiced. If a sound which is normally voiced is, in a particular phonetic environment, produced with less voice than elsewhere, or with no voice at all, it is said to be devoiced. This quality is a matter of degree. In other words, there are contexts in which the devoicing of a particular voiced sound is complete, whereas in other phonetic contexts its voicing is only partially diminished. Be that as it may, there only

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

64

The Production and Classification of Speech Sounds

exists one symbol to indicate devoicing: a small circle [˚] that is placed beneath (under-ring) [d̥] or above (over-ring) [ŋ̊] the consonant symbol. As already noted in Section 1.3.2, an example of devoicing is that which occurs with voiced plosives in word-final positions such as the [g] of tag [tæg̊]. By the same token, voiceless phonemes may show different degrees of vocal fold vibration when occurring next to voiced sounds or in intervocalic positions. This phenomenon is known as voicing and is symbolised with a [ˬ] above or below the phonetic symbol, as in [t] in matter [ˈmæt̬ə]. More details on the devoicing or voicing of RP consonants are offered in section 5.2.2. According to the voice-voiceless distinction the phonemes of RP and PSp can be classified as either voiced or voiceless, as shown in Table 6 below (see also the consonant matrix on the IPA, reproduced as Table 3 in Section 1.4.1). Broadly speaking, voiceless consonants are longer and are articulated with greater muscular effort and breath-force than their voiced counterparts, causing a reduction of the preceding vowels or sonorant consonants while the voiced series do not have such an effect (see Chapter 4 for further details). Now turning to energy of articulation, the fortis/lenis contrast refers to the relatively strong or weak degree of muscular force that a sound is made with. In fortis consonants articulation is stronger and more energetic than in lenis ones. Fortis consonants are voiceless and lenis consonants are not always voiced, since some voicing is lost in initial and final positions, and final consonants are typically almost totally devoiced. Medially – i.e. between vowels or other voiced sounds – lenis consonants have full voicing. When initial in a stressed syllable, fortis plosives /p t k/ have strong aspiration (with a brief puff of air), as in pea [pʰiː], whereas lenis plosives are always unaspirated, as in bib [bɪb] (see § 4.2.1 and 5.2.5). Vowels are shortened before a final fortis consonant, as in beat [biˑt], whereas they have full length before a final lenis consonant, as in bead [biːd]. This phenomenon is known as pre-fortis clipping, which was introduced in Section 2.3.1.4 and will be further discussed in Section 5.2.1. In addition, syllable-final fortis stops often have a reinforcing glottal stop, as in set down [seʔt daʊn], whereas syllable-final lenis stops never have one, as in said /sed/ (see § 5.2.8).

AI 2.5 Voiced and voiceless consonants

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

65

Table 6: Voiced and voiceless consonants in RP and PSp RP

PSp

Voiceless

Voiced

Voiceless

Voiced

ptk

bdg

ptk

bdg

mnŋ

mnɲ r ɾ

fθsʃh

vðzʒ

fθsx

ʧ

ʤ

ʧ

ʝ

w j r (ɹ) l

lʎ

2.3.2.2 Place of articulation The place of articulation (also point of articulation) of a consonant is the point of contact where an obstruction occurs in the VT between an active articulator, i.e. an organ that moves (typically some part of the tongue or the lips), and a passive location or passive articulator, i.e. the target of the articulation, or the place towards which the active articulator moves, whether there is actual contact between them or not. Passive articulators are: the teeth, the gums, and the roof of the mouth comprising alveolar ridge, hard palate and soft palate to the back of the throat. Note that the glottis and epiglottis are movable places of articulation that are not reached by any organs in the mouth. The labels used to describe phonemes according to place of articulation are usually based on the passive articulator. From the front of the mouth towards the back, the places of articulation involved in the production of RP sounds are: (1) bilabial, (2) labio-dental, (3) dental, (4) alveolar, (5) palato-alveolar, (6) palatal, (7) velar and (8) glottal, which, except for (8), are shown in Figure 22 below.17

17 There exist two additional places of articulation that are necessary to describe consonants across the languages of the world: uvular, or sounds articulated with a constriction between the back of the tongue and the uvula (e.g. the uvular trill [R] in French as in rouge ‘red’); and pharyngeal, attributed to sounds articulated with a primary stricture occurring in the pharynx (e.g. the pharyngeal fricatives /ħ/ and /ʕ/ in Somali as in [ʕa:di] ‘normal’, [ħol] ‘cane’, although pharyngeal sounds may also occur in English in disordered speech).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

66

The Production and Classification of Speech Sounds

Figure 22: Seven places of articulation relevant for English

Bilabial articulations involve the lower lip and upper lip, so that bilabial sounds are sounds made by the coming together of both lips. In RP and PSp there are three bilabial sounds: /p b m/. The articulators in labio-dental sounds are the lower lip and the upper front teeth in such a way that the lower lip is actively in contact with the upper teeth, as in the production of RP /f v/ and PSp /f/. The articulators operating in dental sounds are the tip of tongue and the rear of the upper front teeth. Dental consonants are made by the tip of the tongue against the teeth. PSp has three dental consonants /t d θ/, whereas RP has only two /θ ð/, which should be more appropriately identified as “interdental” because in their articulation the tip of the tongue protrudes between the upper and lower teeth. Alveolar, on the other hand, refers to a point of articulation in which the tip/blade of the tongue is in contact with the alveolar ridge, as occurs in the pronunciation of the six alveolar phonemes of RP /t d s z n l/ and the five alveolar consonants of PSp /s n l ɾ r/. In a palato-alveolar (or alveolo-palatal or post-alveolar) place of articulation the articulators are the blade/front of the tongue and the front of the hard palate. The sounds produced in this place require a double movement of the tongue towards the area between the alveolar ridge and hard palate: the blade of the tongue makes contact with the alveolar ridge, while the front of the tongue is raised in the direction of the hard palate. This applies to the five palato-alveolar phonemes found in RP /ʃ ʒ ʧ ʤ r/ and the only palato-alveolar phoneme of PSp /ʧ/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

67

To produce palatal sounds the front of the tongue is in contact with or approaches the hard palate. In RP there is one palatal consonant phoneme, /j/ (the initial sound of yet /jet/), whereas in PSp there are three palatal consonants /ʝ ɲ ʎ/. Other palatalised sounds may be heard as variants of a phoneme (marked with a superscript small J [ʲ]), such as [k] in keep [kʲiːp]) (Section 2.3.2.5). Velar sounds are made by raising the back of the tongue against the soft palate. There are three velar phonemes in RP /k ɡ ŋ/, and also three in PSp /k ɡ x/. Lastly, glottal articulations are made in the larynx, due to the closure or narrowing of the glottis, the aperture between the vocal cords. The audible release of a complete closure at the glottis is known as a glottal stop /ʔ/. Figure 23 locates RP consonant phonemes in their corresponding places of articulation (see also the consonant matrix on the IPA, reproduced as Table 3 in Section 1.4.1).18

Figure 23: Places of articulation for RP consonants

2.3.2.3 Manner of articulation Manner of articulation means how the sound is produced. Articulation involves a stricture, or narrowing of the VT, which affects the airstream. There are three 18 There exist two additional places of articulation that are necessary to describe consonants across the languages of the world: uvular, or sounds articulated with a constriction between the back of the tongue and the uvula (e.g. the uvular trill [R] in French as in rouge ‘red’); and pharyngeal, attributed to sounds articulated with a primary stricture occurring in the pharynx (e.g. the pharyngeal fricatives /ħ/ and /ʕ/ in Somali as in [ʕa:di] ‘normal’, [ħol] ‘cane’, although pharyngeal sounds may also occur in English in disordered speech).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

68

The Production and Classification of Speech Sounds

possible types of stricture, which are, along a scale from greatest to least narrowing, as follows: (1) Complete closure, or occlusion, with complete blocking of airflow; this stricture produces sounds known as stops (plosives and nasals), trills and taps. (2) Close approximation is a narrowing that gives rise to friction, producing sounds known as fricatives or spirants. (3) Open approximation forms no obstruction but changes the shape of the VT, thereby altering the nature of the resonance; this gives rise to the production of (central) approximants and lateral approximants. Note that as in the production of vowels, the airflow is also fully unimpeded. Note that affricate sounds often behave as if they were intermediate between stops and fricatives, but phonetically they are sequences of stop plus fricative. Therefore six possible manners of articulation should be identified: (1) plosives, (2) nasals, (3) fricatives, (4) affricates, (5) central and lateral approximants and (6) taps and trills (see also the consonant matrix on the IPA, reproduced as Table 2 in section 1.4.1). For the articulation of plosives, like RP and PSp /p b t d k g/ (as well as [ʔ] in English), a complete closure is made in the VT that is suddenly released; the air pressure which had built up behind the closure rushes out with an explosive sound, as illustrated in Figure 24 below.

AI 2.6 RP plosives

Figure 24: Manner of articulation for plosive stops

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

69

Nasal sounds are produced with a stricture of complete closure in the oral cavity, but with the soft palate being lowered so that the airflow can escape through the nasal cavity, as represented in Figure 25 below.

AI 2.7 RP nasals

Figure 25: Manner of articulation for nasal stops

The closure may be produced in different places of the VT, which determines the production of different nasal sounds. Depending on the location of the closure, nasals are classified as: bilabial (produced by the lips), as in RP and PSp /m/, alveolar (produced between tip of the tongue and the alveolar ridge), as in RP and PSp /n/, velar (produced by the back of the tongue and the velum), as in RP /ŋ/, or palatal (articulated with the middle or back part of the tongue raised to the hard palate), as in PSp /ɲ/. Fricative sounds are made when two organs come so close together that the air moving between them produces audible friction. There is no complete closure between the organs, simply a stricture or narrowing. In RP there are nine fricatives, /f v θ ð s z ʃ ʒ h/, and in PSp, five /f θ s x ʝ/. An additional distinction is usually made in fricative articulation depending on the shape of the aperture produced to release the airflow, as shown in Figure 26 below (Cruttenden 2014: 198–205).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

70

The Production and Classification of Speech Sounds

AI 2.8 RP fricatives

Figure 26: Palatograms of fricative articulations

Palatogram19 (a) in Figure 26 above shows that for the articulation of RP dental fricatives /θ ð/, the airflow is diffusively released through an opening that is technically termed slit, which means that the upper surface of the tongue is smooth. By contrast, palatograms (b) and (c) show that in the production of both RP alveolar (/s z/) and palato-alveolar /ʃ ʒ/ fricatives, the tongue has a median depression, termed groove, so that the outgoing stream of air is channelled along this central groove, which is quite narrow in the case of the alveolars, but a little broader for the palato-alveolars. Grooved fricatives are collectively known as sibilants, in RP /s z ʃ ʒ/ and /s/ in PSp, because they are produced with much noisier, stronger friction than the slit, dental fricatives. Affricate sounds are produced when the air pressure behind a complete closure in the VT is gradually released: the initial release produces a plosive, but the separation that follows is sufficiently slow to produce audible friction, and there is thus a fricative element in the sound also. However, the duration of the friction is usually not as long as would be the case for an independent fricative sound. In RP only /t/ and /d/ are released in this way, producing /ʧ/ and /ʤ/, respectively; while PSp has only one affricate phoneme /ʧ/. AI 2.9 RP affricates For the articulation of (central) approximants one articulator approaches another in such a way that the space between them is wide enough to allow the airstream 19 A palatogram is a graphic representation of the area of the palate contacted by the tongue that is used in articulatory phonetics to study articulations made against the palate (Crystal 2008: 348).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Articulatory features and classification of phonemes

71

through with no audible friction. In RP there are three central approximant phonemes, /w j r (or ɹ)/, in addition to all vowels and vowel glides. Although the status of [w j] in PSp is a debatable issue, here we shall consider them as allophones of /u/ and /i/ respectively (Navarro Tomás 1966 [1946], 1991 [1918]; Quilis and Fernández 1996).

AI 2.10 RP approximants A lateral (approximant) is made where the air escapes around one or both sides of a closure made in the mouth, as in the various types of /l/ in RP and PSp. Typically this is produced with the centre of the tongue forming a closure with the roof of the mouth but the sides lowered and the air escaping without friction. PSp has an additional palatal lateral phoneme /ʎ/, for the articulation of which there is extensive linguo-palatal contact that is, overall, less posterior than that of /ɲ/. To close this section, a distinction should be made between trills and taps. A trill is a sound made by the rapid percussive action of an active articulator against a passive one. The two types of trill that most frequently occur in languages are alveolar (the tongue-tip striking the alveolar ridge, as in the PSp /r/ of carro ‘cart’) and uvular (the uvula striking the back of the tongue, as in French). A single rapid percussive movement – i.e. one beat of a trill – is termed a tap, as in the PSp /ɾ/ of caro ‘expensive’.

2.3.2.4 Orality Related to manner of articulation, this feature concerns the distinction between oral sounds (produced with a raised velum) and nasal sounds, which are uttered by lowering the soft palate. All consonants are oral except for the three nasal phonemes in RP /m n ŋ/ and PSp /m n ɲ/. However, as already noted for vowels (Section 2.3.1), consonants may have nasalised variants, represented with a [n] [m] when followed by a nasal consonant, as in the case of [t] in catnip [ˈkætⁿnɪp] or [p] in topmost [ˈtopᵐməʊst]. Further details on nasalisation may be found in Sections 2.3.2.5 and 5.2.4.

AI 2.11 Nasal versus oral sounds 2.3.2.5 Secondary articulation The basic production of a speech sound may be modified by means of what is known as secondary articulation. These processes include the following: (1)

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

72

The Production and Classification of Speech Sounds

labialisation, (2) palatalisation, (3) velarisation, (4) glottalisation and (5) nasalisation (Lass 1984; Cohn 1990; Barry 1992; Ladefoged and Maddieson 1996). Labialisation (indicated with a superscript [ʷ]) involves the addition of lip-rounding and the elevation of the tongue back. It is used as a cover term for labialised consonants (i.e. those occurring in the vicinity of rounded vowels, especially consonants preceding them in an accented syllable) such as [kʷ] and [ɫʷ] in cool [kʷuːɫʷ], as well as sequences of the form Cw, as in quantity [kʷwɒntəti] (see § 5.2.3). Palatalisation (symbolised with a superscript [ʲ]), on the other hand, refers to the addition of front tongue raising to hard palate – i.e. the tongue takes on an [i]-like shape with a possible [j] off-glide. This what happens to the u-sounds of words like tune, dune, new, assume, beautiful, which are therefore pronounced [juː] (/tjuːn/, /djuːn/, /njuː/, /əˈsjuːm/, /ˈbjuːtəfl ̩/). As a result, the preceding consonants are represented in narrow transcriptions as palatalised [tʲ dʲ nʲ sʲ bʲ]. Contrast the /m/ in me and more: in the first case it is palatalised [mʲiː], and in the second labialised [mʷɔː]. In addition to the notion of adding an “i-colour”, palatalisation also refers to the process whereby a nonpalatal sounds becomes palatal (see § 5.2.3, 5.3.4). Velarisation means the addition of back-of-the-tongue raising towards the velum – i.e. the tongue takes on an [u]-like shape. This is typical of what is known as dark l, represented as [ɫ], as in still, tell, shall, bull. Glottalisation refers to the addition of a reinforcing glottal stop [ʔ]. The English fortis plosives /p t k ʧ/ are regularly glottalised when syllable-final, as in lipstick [ˈlɪʔpstɪʔk] (see § 5.2.8.2). Finally, nasalisation (marked with [˜]) represents the addition of nasal resonance through lowering the soft palate. In English, vowels preceding nasals are often nasalised as in strong [strɒ̃ŋ], man [mæ̃ n] (see § 5.2 and 5.2.4). Some details on these allophonic variants of phonemes have already been given in Section 2.3.2.2, but a more thorough discussion is presented in Chapters 3 and 4, when dealing with the allophonic variants of each of the RP vowels and consonants, and in Chapter 5 when giving an overview of connected speech phenomena.

2.4 Acoustic features of speech sounds In this section we will look at the acoustic features of speech sounds, using waveforms and spectrograms (SFS/WASP Version 1.54), which represent the size and shape of the VT during their production. There also exist specific computer programs that have been devised to test the computer production and recognition of speech sounds as well as to do speech synthesis, but these different dimensions of speech processing lie well beyond the scope of this manual. For

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Acoustic features of speech sounds

73

further details on such programs and/or speech processing techniques and applications, two good overviews are offered in Ladefoged (1996) and Coleman (2005). Now, if we are to describe speech sounds from an acoustic point of view, broadly it can be said that the larynx is their source and the VT is a system of acoustic filters. In Section 1.2.2, we have seen that the glottal wave is a periodic complex wave (a pulse wave) with different alignments of prominence, or energy peaks, at specific frequencies resulting from VT resonance, known as formants, which represent the strongest (“loudest”) components in the signal with the greatest amplitude and are composed of fundamental frequency (F0) and a range of harmonics. Variations with respect to the direction in which the F0 changes with time (roughly between 60-500Hz) are responsible for intonation (Section 6.4). The filtering function of resonators (nasal cavity and oral cavity) reduce the amplitude of certain ranges of frequency while allowing other frequency bands to pass with very little reduction of amplitude. The output of the resonance system always has the Fo of the glottal wave, while the formants F1, F2, F3 are imposed by the VT, and so there exists a correlation between articulation and formant structure. Thus, the sound waves radiated at the lips and the nostrils are a result of the modifications imposed by this resonating system on the sound waves coming from the larynx, where vocal fold vibration switching on and off triggers phonemic differences and contrasts (degrees of voicing and voicelessness). By way of illustration, the response of the VT (with the tongue in neutral position) is such that it imposes a pattern of certain natural frequency regions (e.g. 500 Hz-1500 Hz-2500 Hz for a VT 17cm in length) and reduces the amplitude of the remaining harmonics of the glottal wave, which results in the articulation of a schwa vowel. The respective peaks of energy (formants F1 – F2 – F3) are retained regardless of variations in Fo of the larynx pulse wave. Further modification of formant structure can be obtained by altering tongue and lip shapes. Rounded and protruded lips, for instance, lengthen the “horizontal” resonance chamber and hence lower its resonating properties, thereby lowering F values. Section 2.4.1 below explains that each vowel sound has a different and unique arrangement of formants, which therefore determine its identifying acoustic characteristics. Likewise, the spectrographic peculiarities of vowel glides and consonants are summarised in Sections 2.4.2 and 2.4.3, respectively.

2.4.1 Vowels All vowels are voiced, as shown by the vertical striations in the spectrograms, and this means that they contain a voicing or voice bar, i.e. a dark band running

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

74

The Production and Classification of Speech Sounds

parallel to the bottom of the spectrogram indicating presence of energy at fundamental frequency associated with voicing, usually in the region of 100–200 Hz depending on speaker-specific characteristics. This can be seen in Figure 27 below, extracted from Katz (2013: 197), which presents the spectographic image of the clause Buy Spot food! produced by a male American speaker (assuming that Spot is hungry).

Figure 27: Reading a spectrogram

Note that in Figure 27 vowels, vowel glides and voiced consonants [b d] do have a voice bar, whereas voiceless sounds [p s t d] do not. Vowel formants, in turn, show up on the spectrogram as dark bands running roughly horizontal with the bottom of the page. Modifications of formant structure can be obtained by altering the shape of the VT, simply by varying the height and backness of the tongue as well as the position of the lips. Furthermore, there exists a correlation between the values of F1 (first formant), the centre frequency of the lowest resonance of the VT, and the articulatory and/or perceptual dimension of vowel height (high vs. low vowels, or close vs. open vowels), on the one hand, and between the backness of the tongue and F2 (second formant), reflecting the place of maximal constriction during the production of the vowel, on the other. The correlation of vowel height (closeness) is inversely proportional to F1 values so that closer vowels have lower F1 values than open vowels and vice versa; whereas front vowels have higher F2 values than back values, lip-rounding having the effect of lowering the overall energy throughout the formants. So, in Figure 27 we can see that the value of F1 is higher for [ɑ], an open vowel, than for [u], a close vowel, while the [ɪ] sound of [aɪ], a front vowel, has a higher F2 value than [ɑ] and [u], two back vowels, but the latter displays an overall lowering of its formant pattern as a result of its lip-rounding.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Acoustic features of speech sounds

75

The adult male formant frequencies for all RP vowels as collected by John Wells around 1960 are presented in Table 7 below, where you can see that: (1) the F1 value for close vowels is around 300 Hz and it rises to 600 and 800 from close-mid to open vowels; (2) F2 values are highest (over 2500 Hz) for front vowels; and (3) F3 values correlate with those of F2. Table 7: Formant frequencies of RP vowels

In addition, there are four other features of vowel production with an acoustic correlate that must be observed in spectrograms. Two of them relate to the relative length of the vowel, pre-fortis clipping and rhythmic clipping (Section 5.2.1), and another two reflect whether the vowel has non-delayed onset to voicing or delayed onset to voicing (Section 5.2.2). In cases of pre-fortis clipping, vowels (especially long ones) have a shorter duration of vocal fold activity as a result of their being followed by voiceless consonants in the same syllable, e.g. [iˑ] in leak [liˑk], which is spectrographically represented by a voice bar and striations with a shorter duration. The same acoustic effect occurs in instances of rhythmic clipping, where (long) vowels are followed by more than one syllable in the same rhythmic unit, like the first vowel of leakage [ˈlĭˑkɪʤ] [ĭˑ], which is even shorter than the vowel of leak. Now compare the spectrograms in Figure 28 below.

Figure 28: Waveforms and spectrograms of “leakage”, “leak”, “lead” and “lee”

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

76

The Production and Classification of Speech Sounds

In Figure 28 you can see that the four instances of [i] differ in length: the sound is shortest [ĭˑ] in leakage (rhythmic and pre-fortis clipping); it is clipped [iˑ] in leak (pre-fortis clipping); it has normal length [iː] in lead before a voiced consonant; and it is extra-long [iːː] in lee, a word that is in an open word-final syllable. Non-delayed onset to voicing, on the other hand, is observed in vowels preceded by syllable-initial voiced consonants, as in bee [biːː], which means that they show vocal fold activity immediately after the release of the consonant, with a voice bar and striations being observed immediately after a short explosion bar. By contrast, if a vowel is preceded by a voiceless consonant, as in pea [pʰiːː], then it has a delayed onset to voicing: this means that vocal fold activity begins a while later, after the release of the consonant, and the voice bar and striations begin a while after the explosion bar, with weak random energy being observed along the frequency axis. The spectrographic representation of nondelayed [i], as in bee [biːː], and delayed [i], as in pea [pʰiːː] [i] are illustrated in Figure 29 below.

Figure 29: Waveforms and spectrograms of “bee” and “pea”

2.4.2 Vowel glides We have seen that in glides the tongue moves in order to produce one vowel quality followed by another, thereby modifying the shape and size of the oral cavity. The tongue movement that takes place during the production of vowel glides is represented spectrographically by a transition in the formant pattern from the first to the second vowel pointing in the direction of which the glide is made, as shown in the production of [aɪ] in Figure 27 above, and the spectrograms of the eight RP diphthongs in Figure 30 below. In some cases the slight bends of formant structures indicate how the speaker has diphthongised the vowel sound in question.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Acoustic features of speech sounds

77

Figure 30: Waveforms and spectrograms of [aɪ, eɪ, ɔɪ, aʊ, əʊ, eə, ɪə, ʊə]

2.4.3 Consonants Consonants can be best spotted in spectrograms at transitions, i.e. the edges of the vowels that are next to the consonants, the time period when the mouth is changing shape between consonant and vowel. We have already seen that voiced consonants have a voice bar and show vertical striations in spectrograms corresponding to vocal fold vibration, whereas no such acoustic cues occur during stop articulation, while in aspiration or frication there is a noise component, as shown in Figure 27 above. Now focusing on the acoustic correlates of place of articulation, bilabial sounds display a weak and diffuse spectrum, and the values of F2 and F3 are comparatively low: the locus of F2 is about 700~1200Hz. Alveolar sounds, in turn, show a diffuse rising spectrum and have an F2 value of approximately 1700~1800Hz, while velar sounds display a compact spectrum in which the second and third formant structures have a common origin, the value of F2 usually being high (about 3000 Hz). Let us analyse the acoustic correlates of manner of articulation. Obstruents (plosives, fricatives and affricates, see § 1.3.2, fn 3) involve a complete or almost complete obstruction to the airflow in the VT, and these different degrees of stoppage have clear acoustic correlates. Thus, the three articulatory phases of plosives [p b t d k g], closure of the articulators, the build-up of air behind them, and the final release, are reflected in spectrograms as a gap in the pattern (the first two stages). Figure 31 below shows that, in voiceless stops [p t k] there is a voiceless silent interval (70–140 ms), which is long if unaspirated, and is followed by a burst if aspirated, in which case the formants may be seen in noise. In voiced stops [b d g], the closure is generally shorter, they have a voice bar during closure and the release burst is weaker and has no aspiration.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

78

The Production and Classification of Speech Sounds

Figure 31: Acoustic phases of plosives in [ɑpʰɑː] and [ɑbɑː]

In fricatives the body of air is forced through a relatively narrow passage involving friction, which acoustically results in a continuous high-frequency noise component (random energy pattern with striations) that is easily identifiable on the spectrogram, as shown in Figure 32: alveolar fricatives [s z] have frequencies concentrated in the high range, at 3600–8000 Hz; the palato-alveolar fricatives [ ʃ ʒ] are somewhat lower, in the range of 2000–7000 Hz; labio-dental [f v] and dentals [θ ð] have similar values (1500–7000 Hz vs 1400–8000 Hz); and the glottal fricative [h] has the lowest values (500–6500 Hz), its spectral pattern being likely to mirror that of the following vowel.

Figure 32: Spectrograms showing the noise patterns of fricatives

In addition, voiceless fricatives [f θ s ʃ h] have noise only, which is much stronger in the case of sibilants [s ʃ ], and they are generally longer than their voiced counterparts [z ʒ], which show weaker noise and voicing bar, although non-sibilant ones [v ð] may have no noise at all. All in all, the friction of voiced fricatives is shorter than that the voiceless series.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Acoustic features of speech sounds

79

The spectrographic representations of affricates [ʤ ʧ ] in Figure 33 below shows the characteristics of their components: a break for the stop combined with a high-frequency noise for the fricative, although the stop transition may have a palatalised component, which is not present in alveolar stops, or alternatively there may be brief intervening alveolar friction [s z] before the fricative component of the affricates [ ʃ ʒ] (Cruttenden 2014: 188–209).

Figure 33: Spectrograms of “ages” [ˈeɪʤɪz] and “h’s” [ˈeɪʧɪz]

Moving on to sonorant consonants (see § 1.3.2, fn 4), including nasals [m n ŋ], liquids [l r] and the approximants [w j], acoustically they behave like vowels (especially in intervocalic positions) in that they exhibit a voicing bar along with formant-like structures. As regards the spectrographic picture of nasals, the manner cues include the absence of an explosion bar, with absence of energy around 1000 Hz, as well as the presence of a low-frequency resonance or “murmur” below 500Hz, with nasal formants at about 250, 2500 and 3250 Mhz. There are also abrupt transitions from and into the neighbouring sounds involving the rapid fall and rise in energy as the nasal is made and released, due to additional nasal cavity resonance that results from lowering the softpalate (velum). The spectral shape of each nasal, particularly in connection with the second and third formant transitions to and from F2 and F3, varies slightly with the place of the obstruction in the VT, as for homorganic plosives: minus transitions for /m/, slight plus transitions for /n/, and plus transitions of F2 and minus transitions of F3 for /ŋ/ (Cruttenden 2014: 209–210). Furthermore, experimental research has shown that a key acoustic feature to distinguish bilabial from

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

80

The Production and Classification of Speech Sounds

alveolar nasals is the relative proportion of energy that is present: 395–770 Hz for bilabials and 1265–2310 Hz for alveolars (Kurowski 1987; Ohde 1994). Spectrograms of /m n ŋ/ in sim, sin and sing are shown in Figure 34 below.

Figure 34: Spectrograms of “sim” [sɪm], “sin” [sɪn] and “sing” [sɪŋ]

The nasalisation of vowels is cued by the presence of a low-frequency resonance and an increase in formant damping, that is, the sound waves get weaker as they progress in time, reducing the vibratory movement and amplitude of the sound. Like nasals, laterals exhibit a discontinuity in formant amplitude from adjacent vowels and also have a vowel-like formant structure with weak amplitude, but the values are in the neighbourhood of 250, 1200 and 2400 Hz. The difference between clear [l] (in syllable-initial positions) and dark or velar [ɫ] (in syllable final positions) lies in the fact that they have respectively the timbre of a close front vowel and a close back vowel: as a result, in [l] the values of F1 (low) and F2 (weak) are quite far apart (around 300 and 1600 Hz), whereas they are close together in [ɫ] (around 400 and 600 Hz). For both realisations of /l/ F3 is high and weak (around 2500 Hz). Transitions from and to vowels tend to be slower than those for nasals. Lastly, the formants of non-lateral approximants show, in general, stronger amplitude than laterals and nasals, and exhibit the corresponding vowels but with lower amplitude and lower F1 (all consonants have lower F1 than vowels), as illustrated in Figure 35. The F1 of [ɹ] is between 120 and 600 Hz, the lower the frequency the greater the lip-rounding, while F2 and particularly F3 display lower values between 700 Hz and 1200 Hz. The tap variant [ɾ], in contrast, has very short closure duration (around 30ms). Turning to the glide consonants, the starting point in both cases is that of [uː] or [iː], i.e. about 240 Hz, but [w] has low F2 (within the range of 360–840 Hz) and weak low F3, whereas [j] has high values for both F2 and F3 (within the range of 2280–3600 Hz).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Acoustic features of speech sounds

81

Figure 35: Spectrograms of [aˈja], [aˈwa], [aˈɹa], [aˈla]

Further reading As in the previous chapter, for an overview of the physics of speech we recommend phonetics texts such as Cruttenden (2014, Chapter 2), Lodge (2009, Chapters 2 and 9), Ashby (2005, Chapter 3; 2011, Chapters 1 to 8), Ogden (2009, Chapter 4), Ashby and Maidment (2005, Chapters 2 to 6), Ball and Rahilly (1999, Chapters 1 to 3), and Laver (1994, Chapter 7). The issues of airstream mechanisms and phonation types are covered at a basic level in a variety of volumes, including Ogden (2009, Chapter 10), Ashby and Maidment (2005, Chapter 7), Ashby (2005, Chapter 6), Ladefoged (2001, Chapter 6), and Ball and Rahilly (1999, Chapters 2 and 4). Slightly more advanced discussion of non-pulmonic consonants can be found in Laver (1994, Chapter 6), and detailed accounts of clicks, ejectives and implosives may be found in Ladefoged (2005, Chapters 1 and 14) and Ladefoged and Maddieson (1996, Chapters 8 and Chapter 3). Accessible accounts of vowel classification, albeit focused on English, are presented in, for instance, Cruttenden (2014, Chapters 4 and 8), Ashby (2011, Chapters 6 and 7), Ashby and Maidment (2005, Chapter 5), and Ladefoged (2001, Chapter 4), whereas Hayward (2000, Chapter 6) provides a detailed analysis of vowel acoustics. Turning to consonants, their classification in VPM terms is addressed in Cruttenden (2014, Chapters 4 and 9), Ashby (2005, Chapters 3 and 5; 2011, Chapter 8), Lodge (2009, Chapter 2), Ashby and Maidment (2005, Chapters 3 and 4), and Ladefoged (2001, Chapter 7). Other advanced readings are Ashby (2006, Chapters 3 to 7), Ladefoged and Maddieson (1996) and Laver (1994, Chapters 8 to 10), among many others. Additional bibliographic recommendations concerning (English and Spanish) vowels and consonants are offered in Chapters 3 and 4, respectively.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

82

The Production and Classification of Speech Sounds

Good introductions to acoustic phonetics may be found in Katz (2013, Chapters 12 to 14), Hewlett and Beck (2006, Chapter 17), Ladefoged (2003), Kent and Read (2002, Chapter 5), and Ball and Rahilly (1999, Chapter 9), whereas Cabrera Abreu and Vizcaíno Ortega (2009) propose an acoustically-based method to explain English phonetics and phonology to SSLE.

Exercises 1. Ingressive sounds are those that are produced on an airstream flowing inward through the mouth or nose. Can you give some examples? Do some languages produce this kind of speech sounds? 2. Voice disorders are those medical conditions that can affect the production of speech. Otolaryngology or ENT (ear, nose, and throat) is the branch of medicine and surgery that specialises in their diagnosis and treatment. Vocal injury is not infrequent. It is claimed that, strictly speaking, only 5% to 10% of the population has a completely normal speech and healthy voice. Can you think which are the main disorders that can affect voice? 3. Speech-Language Pathology professionals (Speech-Language Pathologists (SLPs), or informally, speech therapists) specialise in communication disorders as well as swallowing disorders. What are the main conditions that a speech therapist has to deal with in connection with the production of speech sounds? 4. An articulation disorder involves difficulty in producing sounds. A child with speech sound disorders will be hard to understand, which has serious implications, not only in his/her intelligibility but also in his/her socialisation process. Examine and discuss the following examples of errors children can produce and classify them according to what type of speech disorder you think it represents. 1. 2. 3. 4. 5. 6. 7. 8.

*I go to coo on the buh *I have a wosy wabbit *I am a dood dirl *I saw a wittle wamb *I ree a boo *My poo i boken *I have a wed wadio *The tar is toming

(I go to school on the bus) (I have a rosy rabbit) (I am a good girl) (I saw a little lamb) (I read a book) (My spoon is broken) (I have a red radio) (The car is coming)

5. Speech recognition (SR) is a computer science which translates spoken words into text. Think of applications of this area of study in everyday life.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Acoustic features of speech sounds

83

6. Think of the differences between singing, shouting or speaking quietly, and explain them in phonetic terms. How do rapping or chanting differ from these? 7. Are the phonation and articulation processes independent from each other? If so, what are the consequences? 8. Phonetically speaking, what is the difference between a consonant sound and a vowel? And phonemically? 9. Watch the scene from the Pink Panther again, which was mentioned in Chapter 1, exercise 5. Why do you think would was the word chosen for a close-up of Steve Martin? When dubbed into Spanish, what word should be chosen here in order to achieve the same effect? 10. Look at the sequences in Chapter 1, exercise 6 again. Now explain from a detailed articulatory point of view what might cause the foreignness effect.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:04 PM

Chapter 3

3 Vowels and Vowel Glides 3.1 Introduction In this and the following chapter, the sounds of RP are contrasted with each other as well as with those of Spanish. Accordingly, using the articulatory features (basically tongue shape and height, lip shape and duration) and the acoustic measurements already introduced, this chapter compares and contrasts the vowel systems of RP and PSp. The vowel system of RP consists of twenty sounds, of which twelve are monophthongs (seven short /ɪ e ʌ æ ǝ ɒ ʊ/ and five long /i: ɜː ɑː ɔː uː/) and eight are diphthongs /aɪ eɪ ɔɪ ǝʊ aʊ ɪǝ eǝ ʊǝ/). PSp, on the other hand, has nineteen sounds, five of which are pure vowels /a e i o u/ and fourteen are diphthongs (/ei ̯ ai ̯ oi ̯ au̯ eu̯ ou̯ ja je jo ju wa we wo wi/). In addition, in RP there are five diphthongs + [ə] combinations ([eɪə aɪə ɔɪə əʊə aʊə]) that have no equivalents in the Spanish triphthongs ([wei ̯ wai ̯ jei ̯ jai ̯]), as already noted (see § 1.5, Table 4). The contrastive approach has been adopted for three main reasons. Firstly, it allows us to take into account the interlanguage phonological system of learners, considering both the first language (L1) and the (perception of the) second language (L2) sound systems, as well as additional factors at work such as markedness and other universal phonetic principles (Ellis 1985, 1994). Empirical studies show that, on the one hand, prior language experiences have an impact on the way a language is learned (Flege 1991, 1995, 2003; Kuhl et al. 1992; Kuhl 1993; Best 1994, 1995; Best and Tyler 2007; Boomershine 2013), and, on the other, a better awareness of sound perception may help improve production and understanding (Flege 2003; Flege and MacKay 2004). Focusing on L2 vowel production and perception, studies conducted within the Perceptual Assimilation Model (PAM), for example, have found that the structure of listerners’ vowel space is significantly affected by their L1 vowel inventory and that native-language/second-language perceptual similarity acts as a predictor of difficulties in the discrimination of non-native contrasts (Flege 1991; Best 1994, 1995; Best and Tyler 2007; Boomershine 2013). Likewise, Flege’s (1995, 2003) Speech Learning Model suggests that patterns of learning are predictable from assimilation of L2 sounds to L1 ones, so that vowels that are weakly assimilated into L1 categories should be easier to learn than vowels that are more strongly assimilated. Compatible with these two models is the Native Language Magnet (NLM) observation that sound prototypes have a “perceptual magnet” effect with respect to other sounds, so that sounds in the vicinity of a “prototype” in L1 (e.g. one of the five Spanish vowels) are not as easily discriminated

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

Introduction

85

as are sounds of equivalent auditory difference that are not located near a sound prototype (Kuhl et al. 1992; Kuhl 1993). In what follows it will be shown that, although there are additional factors at work, NLM predictions broadly hold true for SSLE. RP sounds /ʌ æ ɑː/, /e ɜː ə/, /iː ɪ/, /ɔː ɒ/, /uː ʊ/, which resemble the L1 PSp prototypes /a/, /e/, /i/, /o/ and /u/ respectively, are difficult to distinguish from their corresponding prototypes because the latter attract to themselves perceptions of sounds that fall under their scope the way a magnet would, and consequently single category assimilations occur. Likewise, PAM predictions seem to be confirmed in that different types of category assimilations (good exemplar, acceptable but not ideal exemplar, or deviant, as opposed to uncategorisable instances) are influenced by the phonemic inventory of L1, among other factors (Cortés Pomacóndor 1999, 2000; Gallardo del Puerto 2005). Moreover, as duration and the tense-lax distinctions are not contrastive features in Spanish, SSLE are expected to be less sensitive than English natives to such cues (García Lecumberri and Cenoz Iragui 1997; Escudero and Chládková 2010). Secondly, by adopting a contrastive approach that notes the existing differences and similarities between the sound inventories of RP and PSp, we can anticipate and thereby help to circumvent the difficulties that SSLE may encounter when producing and hearing RP sounds, as hypothesised in contrastive or error analysis (Lado 1957). Although learning is in general facilitated where there are equivalent features between L1 and L2, this and the following chapters will show that the equivalences that exist between the sound systems and the suprasegmental features of English and Spanish are rarely exact but mostly similar. As a result, so-called interference or transfer mispronunciations are common in the interlanguage of SSLE, who, assuming a more complete correspondence than exists, carry over Spanish sound patterns in cases where English has in fact no parallel sounds or features. In addition to the L1 influence and speaker-dependent variables mentioned above (anatomic and psychological features and capabilities, age, sex, grammatical correctness, rhythmic adequacy, etc.), there exist two additional factors that may affect the way a sound is perceived and produced which therefore should be considered: phonetic context and orthographic pronunciation. It will be shown that, depending on which phonetic context it occurs in, a sound may have different realisations and, as result, it may be categorised differently (Peperkamp et al. 2006; Altenberg 2005; Quilis and Esgueva 1983; Mann and Repp 1980). For instance, we shall see that SSLE may assimilate RP /dʒ/ in such words as gin to Spanish [ ʝ ] because this language has a voiced palatal fricative word-initially, while in age it can be perceived as [tʃ ] owing to the devoicing of word-final voiced sounds that takes place in English (García Lecumberri and Cenoz 2003).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

86

Vowels and Vowel Glides

Similarly, SSLE’s mispronunciations of unstressed syllables generally result from orthographic pronunciations reflecting the spelling of words ( final *[ˈfaɪnal], silence *[saɪlens], rather than vowel weakening (/ˈfaɪnl ̩/, /ˈsaɪləns/) (Gallardo del Puerto 2005: 416; García Lecumberri and Gallardo 2003). But, notwithstanding the difficulties involved in learning English pronunciation, it is nonetheless true that experimental research suggests that perceptual processes improve gradually as production capabilities are developed, which underscores the dynamic or changing nature of the learner’s interlanguage during L2 acquisition (Flege et al. 1994, 1997; Fox et al. 1995; Allen et al. 1995; Chang et al. 2009). Hence, in order to improve perceptual and production capabilities, Spanish speaking (and all other) EFL learners are advised to practise as much as possible. For this purpose, plenty of practice material is included in this book and the supporting EPSS Multimedia Lab. Our third intention in adopting a contrastive approach is to favour holistic learning. Our aim is that the details of each English sound may be learned all at once and simultaneously compared and contrasted not only with those of Spanish as an L1, but also with each other, on the grounds of both primary and secondary cues. In the case of vowels, for instance, primary cues are articulatory features and F1/F2 targets (Flege et al. 1994), as well as sound-orthography correlations (Chapter 7), whereas formant movement and duration are secondary cues (Fox et al. 1995, Iverson and Evans 2007). In what follows a detailed description is offered of RP vowels (Section 3.2) and vowel sequences including diphthongs and triphthongs (Section 3.3) along nine main parameters: (1) IPA symbols, giving the names of the phonetic symbols. (2) Identification, providing the identification of RP sounds in OBR labels. (3) Position in the CVS, plotting the location of RP and PSp sounds in the CVS. (4) Description, describing the articulation of RP sounds. (5) Environment, noting the main contexts of appearance of RP sounds. (6) Spellings (with transcribed examples), listing the spellings of RP sounds in decreasing frequency and noting the most remarkable exceptions (if any).20 (7) Regional and social variants, mentioning the most important alternative pronunciations in RP, making only sporadic references to other accents that 20 In this subsection, the relevance of silent letters and alternative pronunciations of the same spellings are mentioned in passing, as a more detailed account of the relationship between spelling and pronunciation is offered in Unit 7, where one spelling at a time will be dealt with, together with the various sounds (including absence of sound) that it represents. Incidentally, spellings with and without such as and vs. and , are presented as different spellings because they represent different sets of sounds.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

87

are close to it; whereas the allophonic realisations of RP sounds, i.e. their contextually conditioned phonetic variants, are described separately in Chapter 5. (8) Comparison with Spanish and advice, where the sounds of RP are compared and contrasted with those of PSp and some advice on pronunciation for Spanish learners is provided, including comments about which aspects of RP are essential for intelligibility and which features are only relevant to the imitation of a native-like accent. (9) Further practice, presenting recorded (and transcribed) examples extracted from our Sound Bank. As already explained, the Sound Bank, as well as additional keyed audio exercises and all the listening and support material accompanying this book are available at the companion website EPSS Multimedia Lab.

3.2 A comparison of English and Spanish pure vowels Like other Germanic languages, (RP) English shows a relatively “crowded” vowelspace comprising, as already noted, twelve steady-state or pure vowels that are normally plotted in a trapezium and are usually described in OBR terms as follows: /iː/ /ɪ/ /e/ /ɜː/ /ə/

/ʌ/ /æ/

/ɑː/ /ɒ/ /ɔː/

Close, front, tightly-spread, tense and long (free), as in beans /bi:nz/. Near-close, near-front, slightly-spread, lax and short (checked) (see fn. 7), as in bins /bɪnz/. Between close-mid and open-mid, front, slightly-spread, lax and short (checked), as in sell /sel/. Between close-mid and open-mid, central, neutrally spread, tense and long (free), as in first /fɜːst/. Between close-mid and open-mid (in non-final positions) and open-mid (in final positions), central, neutral lip shape, lax and short (checked), as in alone /əˈləʊn/, doctor /ˈdɒktə/. Between open-mid and open, central, neutral, lax and short (checked), as in drugs /drʌgz/. Between open-mid and open, front, neutral, checked (but longer, in many words, than the other short vowels), neither lax nor tense, as in ham /hæm/. Open, near-back, neutral, tense and long (free), as in bark /bɑːk/. Open, back, slightly-rounded, lax and short (checked), as in dock /dɒk/. Between open-mid and close-mid, back, medium lip-rounding, tense and long (free), as in port /pɔːt/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

88

Vowels and Vowel Glides

/ʊ/

Close-mid, back, closed lip-rounding, lax and short (checked), as in cook /kʊk/. /uː/ Close, back, very closed lip-rounding, tense and long (free), as in roof /ruːf/. By contrast, like most Romance languages, Spanish has a more simple vowel system with only five pure vowels /a e i o u/ that are usually symmetrically arranged in an inverted triangle as there is no front-back opposition at the open position. Table 8 below summarises the similarities and differences that exist between the RP and PSp vowel systems along the OBR parameters that define vocalic quality, examples for each vowel phoneme having been already provided in Table 4 (Section 1.5). Table 8: Classification of the vowels of RP and PSp RP

PSp

Tongue position

Tongue position

Tongue Height Front

Close/high

Central

Tense

Lax

iː

ɪ

Mid

e

Open/low

æ

Tense

ɜː

Back

Lax

Front

Central

Back

Tense

Tense

Tense

Lax

Tense

uː

ʊ

i

u

e

o

ə

ɔː

ʌ

ɑː

ɒ

a

From an articulatory point of view, we can say that English vowels are distributed in both the periphery and the central sections of the CVS. They are distinguished according to the tense-lax and long-short oppositions and generally begin abruptly with a strong vibration of the vocal cords that dies away slowly (Stockwell and Bowen 1965; Flege 1989). In Spanish the opposite happens as vocal cord vibration begins gently and stops brusquely. Besides, SSLE may find it difficult to produce or even perceive the tense-lax and long-short distinctions because these contrasts have no phonemic status in Spanish. All Spanish vowels can be considered to be tense and short, and generally to be closer and pronounced more to the front of the mouth than the English short counterparts, without producing the diphthongisations, or vowel glides to [i], [u] or [ə] (Delattre 1965; Stockwell and Bowen 1965; Monroy Casas 1980, 1981, 2012; Fox et al. 1995). Experimental studies have shown that Spanish vowels have roughly the same length as English short vowels, while English long vowels may be twice as long (García Lecumberri and Elorduy 1994), although in some phonetic contexts

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

89

a reduced long vowel and short vowel may have a similar duration (beat [biˑt] – bid [bɪd]; see § 2.3.1.4). None of the Spanish vowels exactly coincides with the area of articulation of English vowels. As already noted, four RP vowels /iː e ɔː uː/ could be regarded as near equivalents of Spanish /i e o u/, even though English /iː uː/ are more centralised, while /e/ and /ɔː/ are more open and closer than their Spanish counterparts, respectively. In Spanish there also exists an allophone of /a/ occurring in closed syllables and before /u/ (baúl ‘trunk’) which approximates RP /ɑː/ (Navarro Tomás 1991; García Lecumberri and Elorduy 1994). Consequently, SSLE should pay special attention to RP /ə ɜː ɪ ʊ ɒ/, not only because of the potential difficulty involved in their production and perception as Spanish does not have central and lax vowels in its phoneme inventory, but also due to their high frequency of occurrence (especially /ə ɪ/) (Monroy Casas 1980, 1981, 2012; Sedláčková 2010; Helman 2004; Swan and Smith 2001; Finch and Ortiz Lira 1982). Likewise, such contrasts as /iː/ – /ɪ/, /e/ – /æ/, /æ/ – /ʌ/, /ɔː/ – /əʊ/ should also be emphasised, not only because they can prove particularly problematic for SSLE, but also because they are present in all varieties of English native pronunciation (Wells 2005). From an acoustic point of view, English vowels are intrinsically longer (Monroy Casas 1980, 2012)21 and are also generally higher in F2 than their Spanish counterparts, which manifests the already noted tendency for RP vowels to be articulated with a fronted tongue position as compared to PSp vowels (Bradlow 1995; Disner 1983). In addition, while English speakers use three dimensions of spectral information, high-low, front-back and duration, SSLE only use the first two, and so they must be able to not only distinguish a larger number of vowels, but also make use of duration and spectral cues that are not normally relevant for Spanish vowel perception (Quilis 1981; Martínez Celdrán 1996; Morrison 2006).22 Turning to their distribution, Spanish shows a higher proportion of vowels than English (43.49% vs. 39.21%). The predominant vowels in Spanish are the most fronted ones /a e o/, and the central ones in English (33.65% vs. 23.98%), as already remarked (Finch and Ortiz Lira 1982; Cruttenden 2014). In 21 According to Lehiste (1970: 18), there exists a correlation across languages between vocalic quantity and vowel height so that low or open vowels tend to be longer than high or close vowels. 22 Although it is true that in Spanish there may exist length differences between the vowels of such words as azar-azahar (‘chance’ – ‘orange blossom’), pelé-peleé (‘(I) peeled’ – ‘(I) fought’), le-lee (‘to him/her’ – ‘s/he reads’), this difference does not involve a paradigmatic contrast (a choice between two different vowels that triggers a change of meaning), but rather it results from the pronunciation of one or two consecutive instances of the same vowel.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

90

Vowels and Vowel Glides

addition, while all Spanish vowels may occur and are pronounced in all three positions in the word, initial, medial and final, both in unstressed and stressed syllables (Navarro Tomás 1968), English exhibits stress-determined and syllabledetermined constraints on the appearance of certain vowels, the most relevant of such constraints being listed below (Fry 1947; Finch and Ortiz Lira 1982): (1) /ə/ is restricted to unstressed syllables unless as part of a vowel glide (e.g. /ɪə eə ʊə/) (2) Lax vowels, /e æ ɒ ʊ ʌ/, never occur in word-final open syllables, whereas tense vowels and diphthongs can appear in such contexts. (3) Tense vowels and diphthongs can never appear before /ŋ/, while lax vowels show no constraints with regard to codas. (4) /ʊ ʊə/ almost never occur word-initially (apart from oomph /ʊmf/, umlaut /ˈʊmlaʊt/ and Uruguay /ˈjʊərəɡwaɪ/). (5) /iː e ɜː/ very rarely appear in unstressed word-initial syllables. One last difference concerns the sound-spelling correspondences, which, in the case of English, are mentioned in what follows and are further detailed in Chapter 7 in order to help (Spanish) EFL learners build up systematic associations between sounds and spellings. At this stage, suffice it to say that, although both languages use the Latin alphabet, the English alphabet consists of 26 Roman letters, whereas the Spanish one also includes the letter ñ and the digraphs ch and ll, even though the letters k and w do not appear very often in Spanish words (Helman 2004). Confusion can arise when the letters used to represent sounds differ between languages, and this is especially common in the case of English and Spanish vowels. Furthermore, sound-spelling correspondences are much more complex in English than in Spanish: the twelve RP vowels may be represented with seventy regular spellings, apart from another seventy that are less common, while Spanish vowels may correspond to only fourteen different spellings (Finch and Ortiz Lira 1982: 41). Given the relative simplicity of the Spanish vowel-space, for SSLE to be able to produce English vowels correctly, they not only must learn new categories, but also need to be able to “hear” and “see” the differences that distinguish the two sound inventories to avoid confusing them. To ease this task, this book departs from the usual practice of describing each RP vowel individually, based on the tense/lax or long/short distinction and moving counter-clockwise around the CVS. Instead, as represented below in Figure 36, which superimposes the vowel triangle of the five PSp vowels (marked with [○]) and the vowel trapezium of RP vowels (marked with [•]) in the CVS, we classify RP and PSp vowels together into five groups corresponding to the five mappings of perceptual vowel-space (represented with dotted lines) that SSLE tend to make, as already

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

91

noted in the introduction of this chapter (Finch 1982; Alcaraz and Moody 1993; García Lecumberri and Cenoz Iragui 1997; Iverson and Evans 2007; Iverson et al. 2006; Iverson and Evans 2007; Gallardo del Puerto 2005; Escudero and Chládková 2010): (1) Group 1, RP /iː ɪ/ vs. PSp /i/, comprising three vowels situated in the close front section of the CVS. (2) Group 2, RP /e ə ɜː/ vs. PSp /e/, grouping together four vowels that are located in the front or central, between half-close and half-open quadrant of the CVS. (3) Group 3, RP /æ ʌ ɑː/ vs. PSp /a/, bringing together four vowels that are articulated within the front, central or back, between half-open and open section of the CVS. (4) Group 4, RP /ɒ ɔː/ vs. PSp /o/, consisting of three rounded vowels produced in the back, open or just below the half close regions of the CVS. (5) Group 5, RP /ʊ uː/ vs. PSp /u/, relating three rounded vowels produced in the back, close or just above the half close regions of the CVS.

Figure 36: The vowels of RP and PSp in the CVS

3.2.1 Group 1: RP /i: ɪ/ vs. PSp /i/ IPA Symbols i Lower-case I (or Latin small letter I). ɪ Small capital I (or Latin letter small capital I).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

92

Vowels and Vowel Glides

Identification /iː/ Close, front, tightly spread, tense and long (free). /ɪ/ Near-close, near-front, slightly-spread, lax and short (checked). Position in CVS

Figure 37: RP /iː ɪ/ vs. PSp /i/ in the CVS

Description For the articulation of RP /iː/ the front of the tongue is raised just below and just behind the front close (or high) position. The lips are spread and the tongue is tense, with the side rims making a firm contact with the upper molars. The quality is that of a centralised and slightly lowered CV [i ]̞ . RP /ɪ/, on the other hand, is articulated by raising a part of the tongue that is nearer to the centre than the front, to just above the close-mid (or half-close) position. The lips are slightly spread, the tongue is lax (compared with the tension for /iː/), and the side rims make a light contact with the upper molars. The quality is a centralised and closer CV 2 [e] = [ë̝] See the animations and videos under RP /iː/ and /ɪ/ (compared with Spanish /i/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 101–105; Gimson 1994: 97–100; Roach 2005: 16, 20; Collins and Mees 2009: 100; Cruttenden 2014: 111–115). Environment /iː/ occurs in open and closed syllables, but not before /ŋ/, whereas /ɪ/ may occur in all positions of the word, including final open syllables, although note the HappY-tensing process mentioned below.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

93

Spellings Table 9: The spellings of RP /iː ɪ/ /iː/ (in stressed syllables)

/ɪ/ (in stressed or unstressed syllables)

23 see /siː/, cheese /ʧiːz/

is /ɪz/, sit /sɪt/, igloo /ˈɪɡluː/

24 be /biː/, these /ðiːz/

rhythm /ˈrɪðəm/, symbol /ˈsɪmbl ̩/

sea /siː/, beans /biːnz/, leaf /liːf/

25 added /ˈædɪd/, faces /ˈfeɪsɪz/

26 piece /piːs/, believe /bɪˈliːv/

sieve /sɪv/, mischief /ˈmɪsʧɪf/

27 seize /siːz/, key /kiː/, geyser /ˈgiːzə/

orange /ˈɒrɪnʤ/, palace /pælɪs/

ski /skiː/, police /pəˈliːs/

busy /ˈbɪzi/, lettuce /ˈletɪs/

aedile /ˈiːdʌɪl/, aeolic /iːˈɒlɪk/

built /bɪlt/, circuit /ˈsɜːkɪt/

Rare spellings

people /ˈpiːpl ̩/ oecumenic /ˌiːkjuːˈmenɪk/

quay /kiː/

debris /ˈdeɪbriː/, precis/ˈpreɪsiː/

Rare spellings

women /ˈwɪmɪn/

breeches /briːˈʧɪz/ (or /brɪˈʧɪz/)

23 However, if the spelling is followed by in stressed syllables, then its normal pronunciation (NP) is /ɪə/, e.g. beer /bɪə/. 24 /iː/ pronunciation corresponds to the spelling usually in stressed syllables that are followed by silent or other vowel letter, as in eve /iːv/ (complete, legal). 25 Note that the NP of is /-ɪd/ in adjectives (e.g. sacred /ˈseɪkrɪd/, wicked /ˈwɪkɪd/), as well as in the past and past participle of regular verbs ending in /-tɪd/ or /-dɪd/ (e.g. wanted, landed). Similarly, the plural noun inflectional ending is /-ɪz/ (e.g. fishes, faces), only if the singular form ends in /s, z, ʃ, ʧ, dʒ/ final sound (e.g. masses, prizes, fishes, witches, judges). 26 Note that the NP of is /iː/ in medial position, but it is /aɪ/ word-finally: lie /laɪ/, die /daɪ/. 27 The frequent combination is pronounced /siː/, but the NP of the spelling is /eɪ/, as in eight /eit/, weight /weit/, and less likely /aɪ/ as in height /haɪt/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

94

Vowels and Vowel Glides

Note that a shorter variety of /iː/, [i] (transcribed without the length mark [ː]), occurs in unstressed word- and stem-final syllables with the spellings listed below:





happy /ˈhæpi/, city /ˈsɪti/, carrying /ˈkæriɪŋ/ Sunday /ˈsʌndi/ (which can also be pronounced /eɪ/, /ˈsʌndeɪ/) motley /ˈmɒtli/ sweetie /ˈswiːti/, ladies /ˈleɪdiz/ antidote /ˈæntidəʊt/, handicap /ˈhændikæp/ he /hi/, she /ʃi/, we /wi/, me /mi/, be /bi/ Chelsea /ˈʧelsi/, Swansea /ˈswɒnzi/

The first case, involving the pronunciation of word final as [i] rather than [ɪ], illustrates what has been called HappY-tensing (or Y-tensing) (Wells 1982), which originated in southern British accents, but nowadays has become the standard RP pronunciation to such an extent that [ɪ] pronunciations are regarded as somewhat old-fashioned and as a characteristic of Conservative RP (Cruttenden 2014: 331).28 Likewise, [i] may also correspond to , , spellings in or immediately before the following unaccented word endings:











substantiation /səbˌstænʃiˈeɪʃn̩/, creation /kriˈeɪʃn̩/ cereal /ˈsɪəriəl/ Guinean /ˈɡɪniən/ nuclear /ˈnjuːkliə/ media /miːdiə/ genial /ˈʤiːniəl/ Armenian /ɑːˈmiːniən/ deviance /ˈdiːviəns/ substantiate /səbˈstænʃieɪt/ audience /ˈɔːdiəns/ ambient /ˈæmbiənt/ studied /ˈstʌdid/ inferior /ˈɪnˈfɪəriə/ glorious /ˈɡlɔːriəs/ atrium /ˈeɪtriəm/ genius /ˈʤiːniəs/

28 An exception to this HappY-tensing tendency concerns the stem-final of -ly adverbs, which tends to be pronounced either [ɪ] or [ə] (rather than /i/), although the suffix itself does conform to the HappY-tensing tendency, as in the case of e.g. easily /ˈiːzəli/, tidily /ˈtaɪdɪli/)

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

95

Finally, [i] can also be found word-medially, usually before a strong vowel in unstressed syllables with the spellings:

orient /ˈɔːriənt/, periodic /ˌpɪəriˈɒdɪk/, concierge /kɒnsiˈeəʒ/ archaeological /ˌɑːkiəˈlɒʤɪkl ̩/, nuclei /ˈnjuːkliaɪ/, create /kriˈeɪt/

Regional and social variants In the spelling subsection we have already seen that in final unaccented position the contrast between /ɪ/ and /iː/ is neutralised because the general tendency is to pronounce [i], a shorter variant of /iː/. Also common among RP speakers is the diphthongisation of /iː/ so that a slight vowel glide tends to be made from a position near to [ɪ], [ɨɪ], especially in stressed open syllables and in final position. Likewise, as a result of the “Southern diphthong shift”, typical of London, Cockney, Birmingham, the South and Midlands of England, /iː/ may be pronounced as [əɪ]. Turning to /ɪ/, the degree of closeness and centralisation of its variants depends on the accentual force falling upon them and their position in the word. It tends to be pronounced [ə] in some final unaccented endings such as -ate, -em, -ible, -itive, -ily, -ity, as in chocolate /ˈʧɒklət/, system /ˈsɪstəm/, possible /ˈpɒsəbl/, positive /ˈpɒzətɪv/, easily /ˈiːzəli/, and quality /ˈkwɒləti/. In contrast, some other unstressed endings such as -ace, -ess have both /ɪ/ and /ə/ pronuncaitions, as in palace /ˈpæləs/ and useless /ˈjuːsləs/, but the general trend is for /ɪ/ to remain the predominant pronunciation as in: -age /ɪʤ/ village /ˈvɪlɪʤ/, -et /ɪt/ -especially following /k, g, ʧ, ʤ/- pocket /ˈpɒkɪt/, target /ˈtɑːgɪt/, budget /ˈbʌʤɪt/, hatchet /ˈhæʧɪt/ (except -let, -ret> /ət/, e.g. scarlet /ˈskɑːlət/, claret /ˈklærət/), and be- /bɪ/, e.g. become /bɪˈkʌm/). Lastly, where the difference between /ɪ/ and /ə/ reflects a contrast in (grammatical) meaning, RP tends to retain the /ɪ/-/ə/ distinction (e.g. offices /ˈɒfɪsɪz/ versus officers /ˈɒfisəz/), which is also the case of the /ɪt/, /ət/-/eɪt/ pronounciations of the spelling (e.g. estimate (n) /ˈestimət/ vs. estimate (v) /ˈestimeɪt/). The neutralisation of these oppositions suggests a non-RP pronunciation. Comparison with Spanish and advice RP /iː/ is longer and slightly closer than PSp /i/, while RP /ɪ/ has different spectral features: it is not so tense and it is closer and more retracted. SSLE (normally in the initial learning stages) tend to confuse English /i/ and /ɪ/, assimilating both phonemes to the single L1 category /i/. This assimilation derives from a combination of factors: (1) the phonetic similarity and close proximity of these vowels in vowel-space; (2) the fact that Spanish does not have [ɪ] in its inventory, and (3) orthographic reasons since both English /i/ and /ɪ/ may correspond to in spelling, which is the letter that represents Spanish /i/ (García Lecumberri and Cenoz Iragui 1997; Escudero 2000; Escudero and

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

96

Vowels and Vowel Glides

Boersma 2004; Flege et al. 1997; Morrison 2002; Iverson and Evans 2007). However, other SSLE (probably in later learning stages) tend to rate /ɪ/ as closer to Spanish /e/ because of their higher spectral similarity, the strongest cue in their L1, and yet a minority of learners can categorise it as closer to Spanish /u/ (Flege 1991; Escudero 2000; Escudero and Boersma 2004; Morrison 2006, 2008; Boomershine 2013; Escudero and Chládková 2010). Learners are thus advised to practise the RP /iː/ – /ɪ/ contrast in order to be able to perceive and produce their different spectral characteristics (tense vs. lax, long vs. short, closer and more advanced vs. more open and retracted). To produce RP /iː/, SSLE should try to utter a longer sound than their usual Spanish /i/, but the tongue should be in a position somewhat more retracted towards the pharyngeal cavity and the tongue tip should be in contact with the lower incisors; whereas the quality of English [i] (in unstressed syllables and cases of HappY-tensing) is similar to that of Spanish [i]. Turning to RP /ɪ/, to pronounce it Spanish speakers should try to make a shorter sound that is closer to the vowel of the Spanish word mes ‘month’ than to the vowel of mis ‘my’ (Monroy Casas 1980, 2012; Estebas Vilaplana 2009). Further practice AI 3.1. The spellings of RP /iː/. Further practice bee /biː/ believe /bɪˈliːv/ everybody /ˈevribɒdi/

ladies /ˈleɪdiz/ lead /liːd/ leat /liːt/

sheep /ʃiːp/ these /ðiːz/

Please, take these seats. ||ˈpliːz | ˈteɪk ðiːz ˈsiːts|| Jean, can you feel these? ||ˈʤiːn | kən ju ˈfiːl ˈðiːz|| She believes discreet people perceived the scene. ||ʃi bɪˈliːvz dɪˈskriːt ˈpiːpl ̩ pəˈsiːvd ðə siːn|| The queen in green screamed. ||ðə ˈkwiːn ɪn ˈɡriːn ˈskriːmd|| My eel’s meal is a piece of meat with a bit of cheese. ||maɪ ˈiːlz ˈmiːl ɪz ə ˈpiːs əv ˈmiːt wɪð ə bɪt əv ˈʧiːz|| At least, Denise could sneeze and feed and freeze the fleas. ||ət ˈliːst | dəˈniːz kəd ˈsniːz ənd ˈfiːd ənd ˈfriːz ðə ˈfliːz|| Through three cheese trees three free fleas flew. ||θru: ˈθri: ˈtʃi:z ˈtri:z ˈθri: ˈfri: ˈfli:z ˈflu:||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

97

AI 3.2 The spellings of RP /ɪ/. Further practice believe /bɪˈliːv/ illicit /ɪˈlɪsɪt/ lettuce /ˈletɪs/

lid /lɪd/ lit /lɪt/ private /ˈpraɪvɪt/

ship /ʃɪp/ this /ðɪs/

Can you fill this? ||kən ju ˈfɪl ˈðɪs|| When did you begin to build a gym in England? ||ˈwen dɪd ju bɪˈɡɪn tə ˈbɪld ə ʤɪm ɪn ˈɪŋɡlənd|| Although she was busy, she managed to fix the bin in the kitchen. ||ɔːlˈðəʊ ʃi wəz ˈbɪzi | ʃi ˈmænɪʤd tə fɪks ðə bɪn ɪn ðə ˈkɪʧɪn|| Imagining managing an imaginary menagerie. ||ɪˈmæʤɪnɪŋ ˈmænɪʤɪŋ ən ɪˈmæʤɪnəri mɪˈnæʤəri|| Six sick hicks nick six slick bricks with picks and sticks. ||sɪks sɪk hɪks nɪk sɪks slɪk brɪks wɪð pɪks ənd stɪks|| A cricket critic. ||ə ˈkrɪkɪt ˈkrɪtɪk|| 3.2.2 Group 2: RP /e ɜː ə/ vs. PSp /e/ IPA Symbols e Lower-case E (or Latin small letter E). ɜ Reversed Epsilon (or Latin small letter reversed open E). ə Schwa (or Latin small letter Schwa). Identification /e/ Between close-mid and open-mid, front, slightly-spread, lax and short (checked). /ɜː/ Between close-mid and open-mid, central, neutrally spread, tense and long (free). /ə/ Between close-mid and open-mid (in non-final positions) and open-mid (in final positions), central, neutral lip shape, lax and short (checked).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

98

Vowels and Vowel Glides

Position in the CVS

Figure 38: RP /e, ɜː, ə/ vs. PSp /e/

Description For the articulation of RP /e/, the front of the tongue is raised below the closemid position. The lips are slightly spread and slightly wider apart than for /ɪ/. The tongue may have more tension than in the case of /ɪ/, the side rims making a light contact with the upper molars. The quality lies between that of CV 2 [e] and that of CV 3 [ɛ] = [e̞] or [ɛ̝]. RP /ɜː/ is articulated with the centre of the tongue raised between close-mid and open-mid position, the side rims making no firm contact with the upper molars. The lips are neutral. The quality is remote from all peripheral Cardinal Vowel values. Often coinciding in quality with /ɜː/ but differing in length and tension, RP /ə/ is produced with the centre of the tongue being raised between the closemid and open-mid positions in non-final contexts, but in final positions the vowel tends to be articulated in the open-mid central region or even lower [ɐ] (near-open central vowel). The lips are neutral. Like that of /ɜː/, the quality of RP schwa is remote from all peripheral Cardinal Vowel values and its acoustic formants are similar to those for /ɛː/ and /ʌ/. See the animations and videos under RP /e/, /ɜː/ and /ə/ (compared with Spanish /e/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 101–105; Gimson 1994: 101, 115–118; Roach 2005: 16, 20; Collins and Mees 2009: 100; Cruttenden 2014: 111–115).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

99

Environment /e/ is never pronounced in word-final, open syllables. Nearly all cases of /ɜː/ occur in stressed syllables that have an r in the spelling, so care must be taken to avoid post-vocalic /r/ (unless in instances of liaison as described in Section 5.5). Lastly, /ə/ has a very high frequency because it may result from the pronunciation of any vowel or vowel glide in unstressed syllables, as explained in the spelling subsection. However, /ə/ may also occur in stressed syllables as part of a centring glide, /ɪə/, /eə/, /ʊə/ (e.g. dear /dɪə/, where /weə/, tourist /ˈtʊərɪst/), or a triphthong, /aɪə/, /eɪə/, /ɔɪə/, /aʊə/, /əʊə/ (e.g. liar /laɪə/, player /pleɪə/, coyer /kɔɪə/, our /aʊə/, slower /sləʊə/). Spellings In addition to the spellings listed in Table 10, /ə/ is the NP of unstressed function words such as a, an, and, the, for (Section 5.7), and it is also frequent in unstressed prefixes and suffixes, as well as in a large number of unstressed place-names suffixes such as the following:









admonish /ədˈmɒnɪʃ/ Flamborough /ˈflæmbərə/ Edinburgh /ˈedɪnbrə/ Newbury, Newbery /ˈnjuːbəri/ Norfolk /ˈnɔːfək/ Oxford /ˈɒksfəd/ Birmingham /ˈbɜːmɪŋəm/ Ireland /ˈaɪələnd/ establishment /ɪˈstæblɪʃmənt/ Bournemouth /ˈbɔːnməθ/ Maidstone /ˈmeidstən/ succeed /səkˈsiːd/

In contrast, less common or more learned words tend to have a full vowel, unreduced to schwa, in unstressed syllables as in the examples below; while word final (as in tomorrow, yellow) and (as in potato, photo) are normally pronounced [əʊ], schwa reduction being regarded as a very substandard: asphalt /ˈæsfælt/ carnation /kɑːˈneɪʃn̩/ chaos /ˈkeɪɒs/ contract (n) /ˈkɒntrækt/ convert (n.) /ˈkɒnvɜːt/ epoch /ˈiːpɒk/ esquire /esˈkwaɪə/, /ɪˈskwaɪə/ ferment /ˈfɜːment/ (n) (vs. /fəˈment/ (v))

handicap /ˈhændɪkæp/ muscology /mʌsˈkɒləʤi/ Norwegian /nɔːˈwiːʤn̩/ prestige /presˈtiːʒ/, /presˈtiːʤ/ statute /ˈstætjuːt/, /ˈstæʧuːt/ torment /ˈtɔːment/(n) /tɔːˈment/ (v) zodiac /ˈzəʊdiæk/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

100

Vowels and Vowel Glides

Table 10: The spellings of RP /e ɜː ə/ /e/ (in stressed syllables)

/ɜː/ (in stressed syllables)

wet /wet/ pencil /ˈpensl ̩/

herb /hɜːb/ Berkley /ˈbɜːkli/

29 head /hed/, breath /breθ/

err /ɜː/

any /ˈeni/, ate /et/

Thursday /ˈθɜːzdeɪ/

said /sed/ against /əˈɡenst/

purr /pɜː/

says /sez/

first /fɜːst/

Leicester /ˈlestə/

whirr / wɜː/

leopard /ˈlepəd/

myrtle /ˈmɜːtl ̩/

bury /ˈberi/, burial /ˈberɪəl/

myrrh /mɜː/

/ə/ (in unstressed syllables) Any vowel or vowel sequence in unstressed position (esp. , , ) policeman /pəˈliːsmən/ possible /ˈpɒsəbl ̩/ August /ˈɔːɡəst/ chieftain /ˈʧiːftən/ tortoise /ˈtɔːtəs/ ancient /ˈeɪnʃənt/ vengeance /ˈvenʤəns/

Unstressed sequences of vowel(s) plus figure /ˈfɪɡə/ colour /ˈkʌlə/ forget /fəˈɡet/ surpass /səˈpɑːs/

+ world /wɜːld/, worse /wɜːs/, worth /wɜːθ/, worthy /wɜːði/

Unstressed spelling centre /ˈsentə/ genre /ˈʒɑːnrə/

earth /ɜːθ/, earn /ɜːn/

journey /ˈʤɜːni/ French loanwords

milieu /ˈmiːljɜː/ Richelieu /ˈriːʃəˌliзː/

voyeur /vwɑːˈjɜː/ Rare spellings were /wɜː/ colonel /ˈkɜːnl ̩/

29 In the stressed syllables of approximately sixty words because the NP of is /iː/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

101

Lastly, schwa is often lost in many positions (only if leaving an acceptable consonant sequence) as a consequence of the syncopated rhythm of English, as illustrated in the examples below, where you can see that this is quite normal before /l/, /r/ and /n/, which become syllabic. In general, schwa elision and vowel reduction depend a great deal on speed of delivery, so that in rapid speech many reductions occur which would be unusual in slower speech. By contrast, schwa tends to be kept between nasals (Germany /ˈʤɜːməni/) and between a nasal and a homorganic consonant (London /ˈlʌndən/) (see also section 1.3.3.3 on syllabic consonants). aspirin /ˈæspr̩in/ factory /ˈfæktr̩i/ nursery /ˈnɜːsr̩i/ comfortable /ˈkʌmftəbl ̩/ generally /ˈʤenr̩əli/ opening /ˈəʊpn̩ɪŋ/ comparable /ˈkɒmpr̩əbl ̩/ generous /ˈʤenr̩əs/ considerable /kənˈsɪdr̩əbl ̩/

history /ˈhɪstr̩i/ reasonable /ˈriːzn̩əbl ̩/ decorative /ˈdekr̩ətɪv/ literature /ˈlɪtrəʧə/ temperature /ˈtemprəʧə/ definitely /ˈdefn̩ətli/ moderate /ˈmɒdr̩ət/ vegetable, /ˈveʤtəbl ̩/ every /ˈevr̩i/

Regional and social variants Generally RP /e/ tends to have an open quality ([ɛ̝]) that is more similar to CV 3 [ɛ] than to CV 2 [e]; and central diphthongisations of the vowel (e.g. [meən] [weət]) are characteristic of affected RP pronunciation. As regards the other two central vowels, in Refined RP a more open variant of [ɜ̞ː] tends to be produced, while the degree of openness of /ə/ ranges widely depending on the context of appearance. The closest realisation, in the closemid central region of the CVS, occurs in the vicinity of velar consonants /k, g, ŋ/ (e.g. long ago /ˈlɒŋəˈɡəʊ/). A more open variant with tongue-raising between close-mid and open-mid is produced in non-final positions (e.g. alone /əˈləʊn/, fatigue /fəˈtiːɡ/, decorative /ˈdekərətɪv/, afterwards /ˈɑːftəwədz/), while the most open realisation, with the tongue in the open-mid central position, occurs in final positions (e.g. doctor /ˈdɒktə/, mother /ˈmʌðə/, over /ˈəʊvə/, picture /ˈpɪkʧə/). In this context, refined RP speakers tend to produce an even more open [ɐ] and retracted variant [ɑ̙]. In Irish English, many speakers insert schwa between clusters of sonorants, usually /lm/, e.g. /fɪlm/ → /fɪl.əm/. This process leads to resyllabification and is known as schwa epenthesis. For further contexts of schwa pronunciation the reader is referred to Section 5.7 on vowel gradation. Comparison with Spanish and advice Spanish /e/ is much closer to RP /e/ than to /ɜː/, the other e-type vowel of English. RP /e/ is between PSp /e/ and /a/ for F1 but much nearer to the former

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

102

Vowels and Vowel Glides

for F2, and probably for that reason Spanish learners tend to equate it with the Spanish prototype /e/, although less frequently it is also identified with [æ] (García Lecumberri and Cenoz Iragui 1997; Finch and Ortiz Lira 1982). However, despite being quite similar in quantity and despite being transcribed with the same symbol, the quality of Spanish and English /e/ is rather different: RP /e/ is a bit more open and retracted and, to produce it, Spanish learners should try and utter a long Spanish [eː] (especially before voiced consonants) but in the middle the mouth should be opened a bit more by lowering the jaw and the tongue, the tongue tip being in contact with the lower incisors. Its quality is very similar to the Galician e-sound in such words as terra ‘earth’ or tes ‘(you) have’, or the Catalan e-sound in pera ‘pear’ or mel ‘honey’ (Monroy Casas 1980, 2012; Estebas Vilaplana 2009). Turning to /ɜː/ and /ə/, these are the two RP vowels that least resemble any of the PSp vowels, and for this reason they may be difficult for SSLE according to error analysis (Lado 1965). By contrast, the NLM theory hypothesises the opposite: that these two phonemes should be highly distinctive and easy to perceive because there are no similar prototypes in Spanish to be confused with them. Experimental research, however, confirms the error analysis prediction for these two central vowels, as they display a large amount of wrong identifications resulting from assimilations with almost all prototypical vowels of Spanish: some SSLE learners perceive them as similar to [ʌ] -a retracted kind of Spanish [a̙]-, while others assimilate them to a kind of [æ] or to [o], and a minority to a kind of [u] (Finch and Ortiz Lira 1982; García Lecumberri and Cenoz Iragui 1997; Iverson and Evans 2007; Escudero and Chládková 2010). This is probably due to the fact that both the F1 and F2 of the central RP vowels are closest to PSp /a/ (although F1 is also near to PSp /e/ and /o/), in addition to the learners’ tendency to refer to closest sounding and/or similarly spelt L1 vowels. Another important factor involved in this confusion is the fact that Spanish is a syllable-rhythm language whereas English is a stressed-rhythm language, so SSLE have a tendency to pronounce one of the Spanish vowels (when similarly spelt) in each and every syllable (e.g. revelatory */r[e]v[e]ˈl[a]t[o]r[i]/), while English has schwas in most unstressed syllables if they are not smoothed or compressed (e.g. /ˌrevəˈleitəri/) (see also Sections 3.3.3 and 5.7). In order to avoid having a strong foreign accent, it is therefore crucial for EFL learners to pay special attention to the pronunciation and perception of /ɜː/ and /ə/, the latter in particular because it is by far the most frequently used vowel in English (Finch and Ortiz Lira 1982: 79). It has been noted that /ɜː/ and /ə/ are very similar in quality, the only difference being that /ɜː/ is much longer than /ə/. For their articulation, the lips should be neutral and the mouth should be open very little, adopting a mid-centralised tongue position

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

103

slightly higher than for the articulation of Spanish /a/, as in mouth breathing, and then, without moving any of the articulators, the two sounds should be uttered as when thinking aloud (eeehhh), exerting more tension and longer duration in the case of /ɜː/. English /ə/ is very close to the Catalan vowel at the end of such words as hola ‘hallo’ or mare ‘mother’ (Monroy Casas 1980, 2012; Estebas Vilaplana 2009). Further practice AI 3.3. The spellings of RP /e/. Further practice friend /frend/ health /helθ/ led /led/

leopard /ˈlepəd/ let /let/ perfect (vb) /pəˈfekt/

tech /tek/ then /ðen/

Geoffrey Reynolds, my friend, help me bury the dead. ||ˈʤefri ˈrenəldz | maɪ ˈfrend | ˈhelp mi ˈberi ðə ˈded|| The wretch, concentered all in self. ||ðə ˈreʧ ˌkənˈsentәdˈ ɔːl ɪn ˈself|| His head leant on a fence. ||hɪz ˈhed ˈlent ɒn ə ˈfens|| Many an anemone ate an enemy anemone. ||ˈmeni ən əˈneməni et ən ˈenəmi əˈneməni|| Send toast to ten tense stout saints’ ten tall tents. ||ˈsend ˈtəʊst tə ˈten ˈtens ˈstaʊt ˈseɪnts ˈten ˈtɔːl ˈtents|| Seventy seven benevolent elephants. ||ˈseventi ˈsevn̩ bəˈnevələnt ˈelɪfənts||

AI 3.4. The spellings of RP /ɜː/. Further practice attorney /əˈtɜːni/ journal /ˈʤɜːnəl journey /ˈʤɜːni/

serve /sɜːv/ sir /sɜː/ Turk /tɜːk/

word /wɜːd/ worth /wɜːθ/

The nurse prefers the turquoise purse. ||ðə ˈnɜːs prɪˈfɜːz ðə ˈtɜːkwɒɪz ˈpɜːs|| The court adjourned the trial for the murder of the Turkish merchant, which surprised the world. ||ðə ˈkɔːt əˈʤɜːnd ðə ˈtraɪəl fə ðə ˈmɜːdər əv ðə ˈtɜːkɪʃ ˈmɜːʧənt | wɪʧ səˈpraɪzd ðə wɜːld||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

104

Vowels and Vowel Glides

The curly-haired girl’s first birthday is on the third Thursday of this month. ||ðə ˌkɜːliˈheəd ˈɡɜːlz ˌfɜːst ˈbɜːθdeɪ ˈɪz ɒn ðə ˈθɜːd ˈθɜːzdi əv ðɪs ˈmʌnθ|| A turbot’s not a burbot, for a turbot’s a butt, but a burbot’s not. ||ə ˈtɜːbətz nɒt ə ˈbɜːbət | fər ə ˈtɜːbət z ə ˈbʌt | bət ə ˈbɜːbət s ˈnɒt|| How much myrtle would a turtle hurdle if a turtle could hurdle myrtle? ||ˈhaʊ ˈmʌʧ ˈmɜːtl ̩ wʊd ə ˈtɜːtl ̩ ˈhɜːdl ̩ ɪf ə ˈtɜːtl ̩ kəd ˈhɜːdl ̩ ˈmɜːtl ̩||

AI 3.5. The spellings of RP /ə/. Further practice attack /əˈtæk/ banana /bəˈnɑːnə/

Belgium /ˈbelʤəm/ buttock /ˈbʌtək/

question /ˈkwesʧən/ surpass /səˈpɑːs/

See you later! ||ˈsiː jə ˈleɪtə|| Hundreds of students are taking a miserable breakfast. ||ˈhʌndrədz əv ˈstjuːdnŋts ə ˈteɪkɪŋ ə ˈmɪzrəbl ̩ ˈbrekfəst|| There’s a formal proposal to suspend the sentence. ||ðeəz ə ˈfɔːml ̩ prəˈpəʊzl ̩ tə səˈspend ðə ˈsentəns|| The present for my brother is a book about a wizard. ||ðə ˈprezənt fə maɪ ˈbrʌðər ɪz ə bʊk əˈbaʊt ə ˈwɪzəd|| Upper roller, lower roller. ||ˈʌpə ˈrəʊlə | ˈləʊə ˈrəʊlə|| 3.2.3 Group 3: RP /ʌ æ ɑː/ vs. PSp /a/ IPA ʌ æ ɑ a

Symbols Turned V (or Latin small letter turned V). Ash. Script A (or Latin small letter Alpha). Lower-case A (or Latin Small letter A).

Identification /ʌ/ Between open-mid and open, central, neutral, lax and short (checked). /æ/ Between open-mid and open, front, neutral, checked (but longer than the other short vowels in such words as mad, rag, ham, sand), neither lax nor tense. /ɑː/ Open, near-back, neutral, tense and long (free).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

105

Position in the CVS

Figure 39: RP /ʌ, æ, ɑː/ vs. PSp /a/

Description In order to pronounce RP /ʌ/, the jaws are considerably separated and the centre of the tongue is raised just above the fully open position. The lips are neutral and no contact is made between the tongue and the upper molars. Its quality is similar to CV 4 /a/, but slightly centralised and closer: [ä̝] [ɐ]. For the articulation of RP /æ/, on the other hand, the front of the tongue is raised between the open-mid and open positions, with the side rims making a very firm contact with the back upper molars. The lips are neutrally unrounded. The mouth is more open than for /e/. The quality of RP/æ/ is nowadays nearer CV 4 [a] (front open) than CV 3 [ɛ] (front open-mid). RP /ɑː/, in turn, is articulated with the centre and back of the tongue in the fully open position. The lips are neutrally open and no contact is made between the side rims and the upper molars. The quality is that of a centralised CV 5 [ɑ] = [ɑ̎]. Length reduction as a result of pre-fortis and rhythmic clipping is not as noticeable as in the other long vowels (see § 5.2.1). See the animations and videos under RP /ʌ/, /æ/ and /ɑː/ (compared with Spanish /a/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 110–112; Gimson 1994: 102–107; Roach 2005: 16, 20; Collins and Mees 2009: 98–101; Cruttenden 2014: 119–125).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

106

Vowels and Vowel Glides

Table 11: The spellings of RP /ʌ æ ɑː/ /ʌ/ /æ/ (in stressed and some (in stressed syllables) unstressed syllables)

cupboard /ˈkʌbəd/ hiccup /ˈhɪkʌp/ 31 Monday /ˈmʌndeɪ/ + / 32 enough /ɪˈnʌf/ tough /tʌf/ hiccough /ˈhɪkʌp/

/ɑː/ (in stressed syllables)

/ angry /ˈæŋɡri/ gnat /næt/ photographic /fəʊtəˈɡræfɪk/

+ 30 staff /stɑːf/

plaid /plæd/ plait /plæt/

+ can't /kɑːnt/ shan't /ʃɑːnt/

33 blood /blʌd/ flood /flʌd/

+ after /ˈɑːftə/, craft /ˈkrɑːft/

+ demand /dɪˈmɑːnd/ + dance /dɑːns/ + branch /brɑːnʧ/ + gasp /ɡɑːsp/ + last /lɑːst/ + asked /ɑːskt/ + pass /pɑːs/ + path /pɑːθ/ + (silent l) almond /ˈɑːmənd/ + (silent l) half /hɑːf/ 34 car /kɑː/, yard /jɑːd/

clerk /klɑːk/ Derby /ˈdɑːbi/ 35 heart /hɑːt/ , + 36 draught /drɑːft/, laugh /lɑːf/

Rare spellings

does /dʌz/

twopence /ˈtʌpəns/

French loanwords

lingerie /ˈlænʒəri/ timbre /ˈtæmbrə/ meringue /məˈræŋ/

reveille /rɪˈvæli/

French loanwords /wɑː/ repertoire /ˈrepətwɑː/ bazaar /bəˈzɑː/

moustache /məˈstɑːʃ/ camouflage /ˈkæməflɑːʒ/ pyjamas /pəˈʤɑːməz/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

107

Environment /ʌ æ/ do not occur in open-word final syllables, while /ɑː/ does not appear before /ŋ/. Although /ʌ/ normally occurs in stressed syllables, it can also be found in unstressed ones, as in hiccough (/ˈhɪkʌp/), uphold (/ʌpˈhəʊld/) and unfortunate (/ʌnˈfɔːʧənət/). Spellings In addition, attention should be paid to the following exceptions to the most common spellings of RP /ɑː/ noted in Table 11: halfpenny /ˈheɪpeni/ halfpence /heɪpənts/

+

scarce /skeəs/ -wards /wədz/

+

aspect /ˈæspekt/ aspirin /ˈæsprɪn/

+

classic /ˈklæsɪk/ passage /ˈpæsɪʤ/

+

Catherine /ˈkæθrɪn/ gather /ˈgæðə/ swath /swɔːθ/ wrath /rɒθ/

+

Regional and social variants A closer and retracted realisation of /ʌ/, [ʌ̝]̈ , is typical of Refined RP. Also an /ʌ/-/ɒ/ alternation may be found in RP when the stressed spelling is followed by a nasal as in constable /ˈkʌnstəbl/ or /ˈkɒnstəbl/ (also accomplish, comrade). Like /ʌ/, a closer variety of /æ/, [ɛ], is preferred in Refined RP, but it may also be diphthongised to [ɛ ə] so that such words as bad and sad may be pronounced [bɛəd] and [sɛəd]. In addition, it must be noted that this vowel is now generally longer than the other RP short vowels /ɪ, e, ə, ʌ, ɒ, ʊ/. This lengthening effect is (particularly) noticeable before voiced consonants (e.g. bad, bag, can), in which case /æ/ approaches /ɑː/ in length, and even more so in the South of England. Turning to /ɑː/, a back variant of [ɑ̙] is typical of Refined RP, as is the case of [ɑː] pronunciations instead of [æ] in some words such as gymnastic [ʤɪmˈnɑːstɪk] and Atlantic [əˈtlɑːntɪk]. In a number of other words [æ] and [ɑː] pronunciations alternate. But a pronunciation with /ɑː/ seems to be perceived as Refined RP: (1) in tonic position, lather [ˈlɑːðə], [ˈlæðə], masque [mæsk], [mɑːsk], massage

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

108

Vowels and Vowel Glides

[ˈmæsɑːʒ], [məˈsɑːʒ]; (2) in words having the ending -aph (and all its derivates), photograph [ˈfəʊtəɡræf], [ˈfəʊtəɡrɑːf], telegraph [ˈtɛlɪɡræf], [ˈtɛlɪɡrɑːf]; and (3) the words beginning with trans-, transfer [ˈtrænsfə], [ˈtrɑːnsfə]), transform [ˈtrænsfɔːm], [ˈtrɑːnsfɔːm], transmission [trænzˈmɪʃən], [trɑːnzˈmɪʃən], etc. Note, however, that in some of these cases different pronunciations change the meaning of the word: mass /mɑːs/ (religious service ‘misa’), /mæs/ (uncountable noun ‘masa’); ass /ɑːs/ (‘backside’), /æs/ (‘donkey’). Comparison with Spanish and advice The quality of RP /ʌ/ is quite similar to PSp /a/, but it is shorter and more different from it than RP /ɑː/, so it tends to be perceived as a non-prototypical [a] sound by SSLE, who, in some cases, also identify it with [æ] or even with PSp [o] (Finch and Ortiz Lira 1982; García Lecumberri and Cenoz Iragui 1997; Escudero and Chládková 2010). RP /ʌ/ is more centralised and closer than PSp /a/, but we could say that particularly when followed by voiced consonants, as in words such as nothing /ˈnʌθɪŋ/, mother /ˈmʌðə/, borough /ˈbʌrə/, it sounds close to the Spanish [a] sound appearing in such words as nazi, madera ‘wood’ or vara ‘stick’. For the articulation of /æ/, the lower jaw is in a closer position than for Spanish /a/, the tongue is more advanced and the lips are more spread and tense, producing a quality that is half-way between Spanish /a/ and /e/. Tightening of the pharynx may provide useful pronunciation practice. /æ/ is nearest to Sp /a/ in F1 and to PSp /e/ in F2 /a/. But, as the Fl distance between RP /æ/ and PSp /a/ is smaller than that between RP /æ/ and PSp /e/, SSLE tend to assimilate it more with PSp /a/ (García Lecumberri and Cenoz Iragui 1997; Escudero and Chládková 2010). RP /ɑː/ is longer than /ʌ/ and more open, with the lips in a neutral position. Although it is the vowel of this group that comes closest to the Spanish prototype /a/ with regard to formants and in fact sets the closest relationship between

30 In this and the following spellings, is pronounced /ɑː/ in stressed syllables followed by two consonant letters, the first being or silent . 31 is usually pronounced /ʌ/ in the vicinity of . 32 However, the NP of is /aʊ/. 33 But note that the NP of is /uː/. 34 is pronounced /ɑː/ in stressed word-final syllables (whether followed by a consonant or not). 35 The NP of stressed is /ɜː/, and that of is /ɪə/ as in beard /bɪəd/, hear /hɪə/, or less frequently, /eə/ (pear /peə/), or /ɜː/ (heard /hɜːd/). 36 Note that the NP of stressed is /ɒ/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

109

the vowels in the two languages, the tongue is more central and retracted than for PSp /a̙/, and in both cases it is as low as possible (even more so in English), as when yawning, gargling or saying “aaaa” for a throat exploration (Finch and Ortiz Lira 1982; García Lecumberri and Cenoz Iragui 1997; Escudero and Chládková 2010). Note, however, that the backness of /ɑː/ must not reach that of CV 5, since that will sound affected. To avoid misunderstandings and not to confuse RP /ʌ æ ɑː/, SSLE should therefore keep the following in mind (Monroy Casas 1980, 2012; Estebas Vilaplana 2009): (1) /ʌ/ is the shortest and /ɑː/ the longest a-type vowels in English, while /æ/ has an in-between duration, but it is the longest of short RP vowels. (2) RP /ʌ/ is unrounded as well as closer and more advanced than /ɒ/, which is more open and retracted, and rounded. (3) RP /æ/ is more open and advanced than /ʌ/ and its quality is between Spanish /e/ and /a/: for its articulation the mouth must have a similar degree of opening as for Spanish /a/ but the position of the lips is similar to Spanish /e/. Spanish learners should avoid the common mistake of producing [æ] too close to [e]. Special attention should therefore be paid to the vowel distinctions illustrated in words like marry /ˈmæri/, merry /ˈmeri/, Mary /ˈmeəri/, and cup /kʌp/, cap /kæp/ and cop /kɒp/. Galician and Catalan speakers should have no difficulty in producing this sound because they have it in the phonetic inventories of these two languages. (4) /ʌ/ and /ɑː/ are the two English phonemes that come closest to Spanish /a/, their main difference being length: the former is short, and the latter, long. So it is crucial to maintain especially the length difference between /ʌ/-/ɑː/ for the right pronunciation of such minimal pairs as bun – barn; cut – cart; lust – last; must – mast; stuff – staff. Further Practice AI 3.6. The spellings of RP /ʌ/. Further practice another /əˈnʌðə/ blood /blʌd/ doesn’t /ˈdʌznt/

onion /ˈʌnjən/ tuck /tʌk/ tug /tʌɡ/

twopence /ˈtʌpəns/ wonder /ˈwʌndə/

Get rough and unusual stuff in our wonderful companies in Dublin and London. ||ˈɡet ˈrʌf ənd ʌnˈjuːʒʊəl ˈstʌf ɪn ˈaʊə ˈwʌndəfəl ˈkʌmpəniz ɪn ˈdʌblɪn ənd ˈlʌndən||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

110

Vowels and Vowel Glides

The beloved mother and son jumped up just once and had enough fun under the sun on Monday. ||ðə bɪˈlʌvɪd ˈmʌðər ənd sʌn ʤʌmpt ʌp ʤʌst wʌns ənd həd ɪˈnʌf fʌn ˈʌndə ðə sʌn ɒn ˈmʌndeɪ|| Double bubble gum, bubbles double. ||ˈdʌbl ̩ ˈbʌbl ̩ ˌɡʌm | ˈbʌbl ̩z ˈdʌbl ̩|| Mummies make money and a dozen buns and muffins. ||ˈmʌmiz ˈmeɪk ˈmʌni ənd ə ˈdʌzn̩ ˈbʌnz ənd ˈmʌfɪnz|| Rugged rubber baby buggy bumpers. ||ˈrʌɡɪd ˈrʌbə ˈbeɪbi ˌbʌgi ˈbʌmpəz|| Mrs Hunt had a country cut front in the front of her country cut pettycoat. ||ˈmɪsɪz ˈhʌnt həd ə ˌkʌntri ˈkʌt ˈfrʌnt ɪn ðə ˈfrʌnt əv hə ˌkʌntri ˈkʌt ˈpetikəʊt||

AI 3.7. The spellings of RP /æ/. Further practice album /ˈælbəm/ exam /ɪɡˈzæm/ handbag /ˈhændbæɡ/

hat /hæt/ have /hæv/ plaid /plæd/

plait /plæt/ tack /tæk/

Pat is glad she’s passed her exam. ||pæt s ɡlæd ʃiz pɑːst hər ɪɡˈzæm|| As a matter of fact, it was a flash in the pan. ||əz ə ˈmætər əv fækt | ɪt wəz ə flæʃ ɪn ðə pæn|| How can an ant run so fast? ||ˈhaʊ kən ən ˈænt ˈrʌn ˈsəʊ ˈfɑːst|| I gather that the salad on the black mat has apples, carrots, meringue and a sandwich. ||aɪ ˈɡæðə ðət ðə ˈsæləd ɒn ðə ˈblæk ˈmæt hæz ˈæpl ̩z | ˈkærəts | ˈməˈræŋ ənd ə ˈsænwɪʤ|| If you can’t can any candy can, how many candy cans can a candy canner can if he can can candy cans? ||ɪf ju ˈkɑːnt ˈkæn ˈeni ˈkændi ˈkæn | ˈhaʊ ˈmeni ˈkændi ˈkænz kən ə ˈkændi ˈkænə ˈkæn ɪf hi kən ˈkæn ˈkændi ˈkænz||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

111

AI 3.8. The spellings of RP /ɑː/. Further practice asking /ˈɑːskɪŋ/ calm /kɑːm/ clerk /klɑːk/

hard /hɑːd/ heart /hɑːt/ laugh /lɑːf/

marvellous /ˈmɑːvləs/ tar /tɑː/

Ah! Barbara, Martha and Charles are dancing in the bar in the dark while Margaret is playing the guitar. ||ɑː | ˈbɑːbrə | ˈmɑːθə ənd ˈʧɑːlz ə ˈdɑːnsɪŋ ɪn ðə ˈbɑːr ɪn ðə ˈdɑːk waɪl ˈmɑːɡrɪt ɪs ˈpleɪɪŋ ðə ɡɪˈtɑː|| What a funny dancer Barbara is! ||ˈwɒt ə ˈfʌni ˈdɑːnsə ˈbɑːbrə ˈɪz|| Let’s take a fast photograph. ||ˈlets ˈteɪk ə ˌfɑːst ˈfəʊtəɡrɑːf|| We can’t. ||wi ˈkɑːnt|| It’s too dark. ||ɪts ˈtuː ˈdɑːk|| Stay calm! ||ˈsteɪ ˈkɑːm|| No need for alarm. ||ˈnəʊ ˈniːd fər əˈlɑːm|| The giraffe doesn’t harm. ||ðə ʤɪˈrɑːf ˈdʌznt ˈhɑːm|| 3.2.4 Group 4: RP /ɔː ɒ/ vs. PSp /o/ IPA ɒ ɔ o

Symbols Turned script A (or Latin small letter turned alpha). Latin small letter open O. Latin small letter O.

Identification /ɔː/ Between open-mid and close-mid, back, medium lip rounding, tense and long (free). /ɒ/ Open, back, slightly-rounded, lax and short (checked).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

112

Vowels and Vowel Glides

Location in the CVS

Figure 40: RP /ɒ ɔː/ vs. PSp /o/

Description For the articulation of RP /ɔː/ the back of the tongue is raised between the openmid and close-mid position and no contact is made between the tongue and the upper molars. There is medium lip-rounding. Its quality lies between CV 6 [ɔ] and CV 7 [o], i.e. [ɔ̞] or [o̞]. RP /ɒ/ is articulated with the back of the tongue in the fully open position. The lips are slightly rounded and no contact is made between the tongue and the upper molars. The quality is that of an open lip-rounded CV 5 [ɑ] = secondary CV 13 [ɒ]. See the animations and videos under RP /ɔː/ and / ɒ / (compared with Spanish /o/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 114–118; Gimson 1994: 108–111; Roach 2005: 16, 20; Collins and Mees 2009: 98–101; Cruttenden 2014: 99, 102). Environment /ɒ/ never occurs word-finally, and /ɔː/ does not appear before /ŋ/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

113

Spellings Table 12: The spellings of RP /ɔː ɒ/ /ɔː/ (in stressed syllables)

/ɒ/ (in stressed and unstressed syllables)

war /wɔː/, quart /kwɔːt/ toward /təˈwɔːd/

37 cod /kɒd/, dock /ˈdɒk/, response /rɪˈspɒns/

sword /sɔːd/ forward /ˈfɔːwəd/

// + 38 was /wɒz/, what /ˈwɒt/ quarrel /ˈkwɒrəl/

sore /sɔː/, sure /ʃɔː/

, cough /kɒf/, knowledge /ˈnɒlɪdʒ/

, , 39 four /fɔː/, board /bɔːd/ poor /pɔː/

because /bɪˈkɒz/, cauliflower /ˈkɒlɪflaʊə/

, , + silent awful /ˈɔːfəl/, cause /kɔːz/ daughter /ˈdɔːtə/ + ////40 all /ɔːl/, salt /sɔːlt/ stalk /ˈstɔːk/, also /ˈɔːlsəʊ/ + silent 41 ought /ˈɔːt/, fought /ˈfɔːt/

broad /brɔːd/, abroad /əˈbrɔːd/ Rare spellings drawer /drɔː/ (‘sliding box’) drawer /drɔːə/ (‘person who draws’) awe /ɔː/

Rare spellings + silent yacht /jɒt/ bureaucracy /bjʊəˈrɒkrəsi/

37 Stressed is pronounced /ɒ/ in the following cases: (1) monosyllabic words ending in one or two consonants; (2) words ending in two consonants + silent (e.g. response [rɪˈspɒns], involve [ɪnˈvɒlv]); (3) two-syllable words in which stressed is followed by two consonants or more (e.g. doctor [ˈdɒktə], constant [ˈkɒnstənt], across [əˈkrɒs]) (4) polysyllabic words with the primary stress on (e.g. crocodile [ˈkrɒkədaɪl], dominate [ˈdɒmɪneɪt], positive [ˈpɒzɪtɪv], poverty [ˈpɒvəti], solitary [ˈsɒlɪtəri]); and (5) words ending in , (e.g. comic [ˈkɒmɪk], methodical [məˈθɒdɪkl]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

114

Vowels and Vowel Glides

The following common English words, however, do not conform to the spelling tendencies just described for RP /ɔː/: shall /ʃæl/ (strong form) draught /drɑːft/ laugh /lɑːf/ Regional and social variants With respect to /ɔː/, as noted in fn 36, this phoneme used to contrast with /ɔə/, normally in words derived from in RP, so that saw and sore were pronounced differently. Nowadays this contrast is generally not made, except by some older speakers. In addition, a number of words which formerly had only /ʊə/ in RP have now acquired an alternative pronunciation with /ɔː/ (e.g. sure, poor, your). Note also the SSBE tendency for /ɔː/ to cover both /ɔː/ and /ɒ/, so that, for example, the words caught and cot may be pronounced [kɔːt]. In addition, alternative /ɔː/-/ɑː/ pronunciations are observed in some French loanwords such as launch [lɔːnʃ ], [lɑːnʃ ] and jaundice [ˈʤɔːndɪs], [ˈʤɑndɪs]. /ɒ/, in turn, has very little variation in RP (and generally in British English), apart from the preference of some Refined RP speakers to pronounce [ɔː] instead of [ɒ] in a number of words, usually preceding consonant clusters headed by /l, s, θ, f/ (e.g. across [əˈkrɔːs], [əˈkrɒs], Australia [ɔːsˈtreɪlɪə], [ɒsˈtreɪlɪə], Austria [ˈɔːstrɪə], [ˈɒstrɪə], austerity [ɔːˈstɛrᵻti], [ɒˈstɛrᵻti], auspice [ˈɔːspɪs], [ˈɒspɪs], authority [ɔːˈθɒrᵻti], [ɒˈθɒrᵻti], cloth [klɔːθ], [klɒθ], off [ɔːf], [ɒf], salt [sɔːlt], [sɒlt], false [fɔːls], [fɒls]). Comparison with Spanish and advice RP /ɒ/ and /ɔː/ show acoustic features that place them in the region of Spanish /o/, although the English sounds involve a more retracted realisation, /ɒ/ being more open than /o/, and /ɔː/ closer. Some SSLE, however, may rate [ɒ] as a type of PSp [e] probably due to their close F2 value, and in some cases [ɒ] and [ɔː] may be confused as a result of the inability to perceive the lax-tense and length

38 /ɒ/ is the NP of , , in stressed syllables and when these combinations occur before one or more final non-velar sounds. Nevertheless, the letter has its NP /eɪ/ when these spellings precede: (1) a single consonant + silent , (2) a single consonant + a pronounced vowel, or (3) a velar consonant /k, g, ŋ/ (e.g. whale [weɪl], wage [weɪdʒ], equator [ɪˈkweɪtə], quake [kweɪk]). 39 Formerly this spelling was pronounced [ɔə], but today this is regarded as conservative pronunciation in regional accents. 40 Also in compounds and derivatives in which is single. 41 In stressed syllables of around 30 words; this spelling has many different pronunciations.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

115

cues that distinguish these two sounds (Finch and Ortiz Lira 1982; García Lecumberri and Cenoz Iragui 1997; Escudero and Chládková 2010). Accordingly, for the articulation of these two rounded vowels learners are advised to pay special attention to the perception and production of the lax-tense and short-long contrasts, in addition to their different degrees of lip-rounding, which seems to be the crucial feature to distinguish the three sounds (Monroy Casas 1980, 2012; Ward 1972: 96). To practice RP /ɔː/, as in call /kɔːl/, SSLE may try to produce the vowel of col ‘cabbage’ but with longer duration; its quality is close to that of Galician or Catalan [o] in final positions. To target RP /ɒ/, on the other hand, a good practice may be to utter a long Spanish [o] but in the middle the mouth should be opened a bit more by lowering the jaw and the tongue (Monroy Casas 1980, 2012; Estebas Vilaplana 2009). Remember that RP /ɒ/ is shorter and more open than RP /ɔː/, while Spanish /o/ has an in-between quality. Further Practice AI 3.9. The spellings of RP /ɔː/. Further practice always /ˈɔːlweɪz/ author /ˈɔːθə/ bald /bɔːld/

before /bɪˈfɔː/ law /lɔː/ sought /ˈsɔːt/

sword /sɔːd/ water /ˈwɔːtə/

Sean’s daughter, Laura, adores small tortoises and football. ||ˈʃɒnz ˈdɔːtə | ˈlɔːrə | əˈdɔːz ˈsmɔːl ˈtɔːtəsɪz ənd ˈfʊtbɔːl||/ Paul thought going to Malta in August or in autumn is a bore. ||ˈpɔːl ˈθɔːt ˈɡəʊɪŋ tə ˈmɔːltə ɪn ˈɔːɡəst ɔːr ɪn ˈɔːtəm ˈɪz ə ˈbɔː|| The author of the story, George, has already gone in the yacht for a fourteen minute sail in the cold water. ||ði ˈɔːθər əv ðə ˈstɔːri | ˈʤɔːʤ | həz ɔːlˈredi ˈɡɒn ɪn ðə ˈjɒt fər ə ˌfɔːtiːn ˌmɪnɪt ˈseɪl ɪn ðə ˈkəʊld ˈwɔːtə|| Knife and a fork, bottle and a cork that is the way you spell New York. ||naɪf ənd ə fɔːk | ˈbɒtl ̩ ənd ə ˈkɔːk ˈðæt ɪs ðə ˈweɪ ju ˈspel ˌnjuː ˈjɔːk|| Does this shop sport cords and short socks with spots? ||dəz ðɪs ˈʃɒp ˈspɔːt ˌkɔːdz ənd ˌʃɔːt ˈsɒks wɪð ˈspɒts|| I saw a saw in Warsaw that could outsaw any saw that I ever saw. ||ˈaɪ ˈsɔː ə ˈsɔː ɪn ˈwɔːsɔː ðət kəd ˌaʊtˈsɔː ˈeni ˈsɔː ðət ˈaɪ ˈevə ˈsɔː||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

116

Vowels and Vowel Glides

AI 3.10. The spellings of RP /ɒ/. Further practice because /bɪˈkɒz/ cough /kɒf/ donkey /ˈdɒŋki/

knowledge /ˈnɒlɪʤ/ revolver /rɪˈvɒlvə/ rock /ˈrɒk/

tog /ˈtɒɡ/ wander /ˈwɒndə/

Tommy and Polly wandered around Gloucester and Oxford. ||ˈtɒmi ənd ˈpɒli ˈwɒndəd əˈraʊnd ˈɡlɒstər ənd ˈɒksfəd||/ Sorry, what’s wrong with washing socks? ||ˈsɒri | ˈwɒts ˈrɒŋ wɪð ˈwɒʃɪŋ ˈsɒks||/ Everybody would want a long holiday in Australia. ||ˈevrɪbɒdi wʊd ˈwɒnt ə ˌlɒŋ ˈhɒlədeɪ ɪn ɒsˈtreɪliə||/ If pros and cons are opposite, is progress the opposite of congress? ||ɪf ˈprəʊz ənd ˈkɒnz ər ˈɒpəzɪt | ˈɪz ˈprəʊɡres ði ˈɒpəzɪt əv ˈkɒŋɡres|| Top chopsticks shops stock top chopsticks. ||ˌtɒp ˌʧɒpstɪks ˈʃɒps ˈstɒk ˌtɒp ˈʧɒpstɪks||/ 3.2.5 Group 5: RP /uː ʊ/ vs. PSp /u/ IPA Symbols u Latin small letter U. ʊ Upsilon (or Greek small letter Upsilon). Identification /uː/ Close, back, very closed lip rounding, tense and long (free). /ʊ/ Close-mid, back, closed lip rounding, lax and short (checked). Location in the CVS

Figure 41: RP /ʊ, uː/ vs. PSp /u/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

117

Description For the articulation of RP /uː/ a part of the tongue somewhat centralised from true back is raised just below closed position, in a symmetrical back relationship with the front vowel /iː/. The articulation is tense compared to that of /ʊ/ and no firm contact is made between the tongue and the upper molars. The lips tend to be very closely rounded or trumpet-shaped. The quality is that of a slightly lowered and centralised CV 8 [u] = [ü̞]. RP /ʊ/ is articulated with a part of the tongue nearer the centre than the back raised just above the close-mid position, in a symmetrical back relationship with the front vowel /ɪ/. There is close lip-rounding. The tongue is lax (compared with the tenser /uː/) and no firm contact is made between the side rims and the upper molars. The quality is a centralised and closer CV 7 [o] with closer rip-rounding = [ö̞]. See the animations and videos under RP /uː/ and /ʊ/ (compared with Spanish /u/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 119–122; Gimson 1994: 112–114; Roach 2005: 16, 20; Collins and Mees 2009: 98–101; Cruttenden 2014: 130–134). Environment [ʊ] is found in both accented and unaccented syllables, but it does not occur in word‑initial positions nor before final /ŋ/, and finally only in the unaccented form of to [tʊ] and you [ jʊ]. RP /uː/ is never pronounced before /ŋ/, either. Spellings Note that a shorter variety of [u] (without the length marks [ː]) occurs in unstressed syllables, especially after [j] (circular /ˈsɜːkjulə/), [tj] when it results from the merging of [t] and [j] (situation /sɪtjuˈeɪʃən/ – /sɪʧuˈeɪʃən/, as well as in unstressed word- and stem-final spellings listed below.

you /ju/, throughout /θruˈaʊt/ to /tu/, into /ˈɪntu/, who /huː/

[u] can also occur word-medially, usually before a stressed vowel with the spellings:

interviewee /ˌɪntəvjuˈiː/ influenza /ˌɪnfluˈenzə/, virtuosity /ˌvɜːʧuˈɒsəti/ genuine /ˈʤenjuɪn/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

118

Vowels and Vowel Glides

Table 13: The spellings of RP /uː ʊ/ /uː/ (mostly in stressed syllables)

/ʊ/ (in stressed and unstressed syllables)

42

put /ˈpʊt/, cushion /ˈkʊʃən/, sugar /ˈʃʊɡə/ crucial /kru:ʃəl/, flute /fluːt/ fluency /ˈfluːənsi/, fluor /fluːə/ 43 too /tuː/, moon /muːn/

look /lʊk/, wool /wʊl/

do /duː/, move /muːv/

woman /ˈwʊmən/ Wolverhampton /ˈwʊlvəˌhæmptən/

44 group /gruːp/, wound /wuːnd/ could /kʊd/, should /ʃʊd/ through /θruː/ , , 45 flew /fluː/, blue /bluː/ virtue /ˈvɜːʧuː/, suit /suːt/

canoe /kəˈnuː/

Finally, [u] / [ju] / [jʊ] pronunciations may also be heard in the spellings of such word endings as:





jaguars /ˈʤæɡjuəz/ casual /ˈkæʒjuəl/, actually /ˈækʧuəli/ confluence /ˈkɒnfluənts/ influent /ˈɪnfluənt/, continuance /kənˈtɪnjuəns/ February /ˈfebruəri/ – /ˈfebjʊəri/, mortuary /ˈmɔːʧuəri/ insinuate /ɪnˈsɪnjueɪt/, situation /ˌsɪtjuˈeɪʃn̩/

42 is pronounced /ʊ/ in the stressed syllables of about 30 words of common use. 43 is pronounced /uː/ in stressed syllables, except when followed by , in which case its NP is /ɔː/. 44 The NP of is /aʊ/. 45 The NP of the spellings , , , , , is /uː/ in stressed syllables, and normally after /tʃ dʒ r l/). Note, however, that these spellings are pronounced [juː] when they are preceded by: (1) plosives /b d p t k/ (e.g. tune [tjuːn], duke [djuːk], argue [ˈɑːɡjuː], beauty [ˈbjuːti], askew [əˈskjuː]; (2) nasals /m n/ (e.g. music [ˈmjuːzɪk], neuter [ˈnjuːtə], new [njuː], nuisance [ˈnjuːsns]); (3) fricatives /f v h/ (e.g. feud [fjuːd], fuel [ˈfjuːəl], few [fjuː], hue [hjuː]); and (4) , if preceded by a stressed vowel (e.g. value [ˈvaljuː]). But both /uː/ and /juː/ are heard when the spellings , , , are preceded by: (1) /s z θ/ (e.g. enthusiasm [enˈθuːzɪæz(ə)m], [enˈθjuːzɪæz(ə)m], suit [suːt], [sjuːt]), and (2) syllable-initial /l/ or when /l/ is preceded by an unstressed vowel (e.g. absolute [ˈabsəluːt], [ˈabsəljuːt], lute [luːt], [ljuːt]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish pure vowels

119

Regional and social variants Apart from variation in rounding and variation between /ʊ/ and /uː/ in some words (e.g. broom [brʊm], [bruːm], groom [ɡrʊm], [ɡruːm], room [rʊm], [ruːm], tooth [tʊθ], [tuːθ]), little striking variety is found in RP /ʊ/. There exists an increasing tendency for [ʊ] to be unrounded in such common words as e.g. could [kʊd], should [ ʃʊd], would [wʊd], which have a schwa in their unaccented forms, so if we take into account its unrounding process, it is coming closer to its unrounded equivalent, that is secondary CV 15 [ɤ]. Throughout the north of England no contrast is made between /ʊ/ and /ʌ/, and a vowe1 in the region of [ʊ] is used for both. Turning to RP /uː/, there are two main variants: (1) more centralised realisations or fronted variants if following [j] (e.g. beauty [ˈbjüˑti], youth [jüːθ); and (2) different kinds of diphthongisation [ʊu], [əʊ], particularly in final position (e.g. do [dʊu], [dəʊ], shoe [ ʃʊu], [ ʃəʊ], who [hʊu], [həʊ]) in Cockney. In SSE, as there is no /ʊ/-/uː/ contrast, both vowels are covered by a centralised kind of CV8 [u] = [ü], or by a centralised vowel with a very slight lip-rounding [ÿ]. Comparison with Spanish and advice Both RP /uː/ and /ʊ/ tend to be identified with PSp /u/, which make it hard for Spanish speakers to be able to distinguish such English words as fool /fuːl/ and full /fʊl/ (Finch and Ortiz Lira 1982; García Lecumberri and Cenoz Iragui 1997; Escudero and Chládková 2010). However, the quality and duration of both RP sounds differ from their Spanish counterpart. /uː/ is longer and has more tension and correspondingly more lip-rounding and a closer realisation than PSp /u/. To produce English /uː/, Spanish speakers should utter a lengthened Spanish [u], as the vowel of the word uva ‘grape’, whereas the quality of English [u] (in unstressed syllables) is similar to that of Spanish [u]. RP /ʊ/, in turn, is shorter than /uː/ and more open than PSp /u/, its quality lies between Spanish [u] and [e]. For its pronunciation a good practice may be to utter a long Spanish [u], but in the middle the mouth should be opened a bit and the lips should be relaxed so as to have a bit less lip rounding. As /ʊ/ is half-way between RP /u/ and /ə/, any error in its quality should be in the direction of /ə/ (Monroy Casas 1980, 2012; Estebas Vilaplana 2009). Further Practice AI 3.11. The spellings of RP /uː/. Further practice afternoon /ˌɑːftəˈnuːn/ lose /luːz/ supernatural /ˌsuːpəˈnæʧrəl/ flu /fluː/ rheumatism /ˈruːmətɪzəm/ tissue /ˈtɪʃuː/ loose /luːs/ suit /suːt/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

120

Vowels and Vowel Glides

Hugh and Sue were truly truants and liked cool, good food. ||ˈhjuː ənd ˈsuː wə ˈtruːli ˈtruːənts ənd ˈlaɪkt ˈkuːl | ˈɡʊd ˈfuːd|| Excuse me, Andrew, do you have chewing gum and some fruit juice? ||ɪkˈskjuːz ˈmiː | ˈændruː | də ju ˈhæv ˈʧuːɪŋ ˌɡʌm ənd səm ˌfruːt ˈʤuːs|| My wounds were oozing pus and I had to reschedule the junior troop meeting. ||ˈmaɪ ˈwuːndz wər ˈuːzɪŋ ˈpʌs ənd ˈaɪ ˈhæd tə ˌriːˈʃedjuːl ðə ˈʤuːnɪə truːp ˈmiːtɪŋ|| June assumes that by looking at the manual she’ll produce a solution with a blue coloration. ||ˈʤuːn əˈsjuːmz ðət ˈbaɪ ˈlʊkɪŋ ət ðə ˈmænjʊəl ʃil prəˈdjuːs ə səˈluːʃn̩ wɪð ə ˌbluː ˌkʌləˈreɪʃn̩|| A Tudor who tooted the flute tried to tutor two tooters to toot. ||ə ˈtjuːdə huˈtuːtɪd ðə ˈfluːt ˈtraɪd tə ˈtjuːtə ˈtuː ˈtuːtəz tə tuːt||

AI 3.12. The spellings of RP /ʊ/. Further practice butcher /ˈbʊʧə/ neighbourhood /ˈneɪbəhʊd/ pudding /ˈpʊdɪŋ/

Roumania /rʊˈmeɪniə/ sugar /ˈʃʊɡə/ wolf /wʊlf/

wools /wʊlz/ worsted /ˈwʊstɪd/

Please, put the cookery books on the bookshelf in the dining-room, would you? ||pli:z | ˈpʊt ðə ˈkʊkəri ˌbʊks ɒn ðə ˈbʊkʃelf ɪn ðə ˈdaɪnɪŋ ˌru:m | wʊd ˈju:|| The woman from Worcester suffered from bosom related problems. ||ðə ˈwʊmən frəm ˈwʊstə ˈsʌfəd frəm ˌbʊzəm rɪˈleɪtɪd ˈprɒbləmz|| Hugh first stood looking at the books, and then pulled the covers of Puss in Boots, threw them to the brook and told the truth, without beating about the bush. ||ˈhju: ˈfɜːst ˈstʊd ˈlʊkɪŋ ət ðə ˈbʊks | ənd ˈðen ˈpʊld ðə ˈkʌvəz əv ˈpʊs ɪn ˈbu:ts | ˈθru: ðəm tə ðə ˈbrʊk ənd ˈtəʊld ðə ˈtru:θ | wɪðˈaʊt ˈbiːtɪŋ əˈbaʊt ðə ˈbʊʃ|| Esau Wood saw a wood saw, saw wood, as no wood saw would saw wood. ||ˈɪsɔː ˈwʊd ˈsɔː ə ˌwʊd ˈsɔː | ˈsɔː ˈwʊd | əz ˈnəʊ ˌwʊd ˈsɔː wʊd ˈsɔː ˈwʊd||

3.3 A comparison of English and Spanish vowel glides Glides are articulatorily described as movements from a given vocalic starting point towards an end-point alluding to two main parameters: (1) Tongue shape, including tongue backness and height, which refers to which part of the tongue is raised and how close it comes to the roof of the mouth (e.g. (just, slightly) above, below, behind, forward of, front/centre/ back, (half) open/close position);

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

121

(2) Lip shape, which generally changes during glides in accordance with the lip spreading/rounding associated with the vowel elements involved, though the amount of lip shape change is often less than might be expected. Accordingly, under the term diphthong are included sequences of two vocalic elements forming a movement or glide within one syllable from a first element, or starting point, towards a second element, or end-point, as opposed to pure vowels which remain constant and do not glide. Diphthongs are usually classified articulatorily and auditorily according to three main parameters: (1) The direction of the movement of the tongue, distinguishing closing (if the glide is in the direction of a closer vocalic position), opening (if the movements is towards a more open position) or centring diphthongs (if there is a glide from a peripheral to a more centralised vocalic position); (2) The distance the tongue travels, according to which diphthongs can be articulatorily labelled narrow (if the glide is short) and wide (if the glide is long); (3) The prominence of their elements, which determines whether diphthongs are auditorily falling (when most of the length and stress associated with the glide is concentrated on the first element, the second element being only slightly sounded) or rising (when the prominence increases as we pass from the first to the second element). Focusing on English and Spanish, the diphthongal space shows the reverse situation than that found for vowels because RP has fewer diphthongs than PSp, as already noted in the introduction of this chapter (see also § 1.5 Table 4). RP has only eight diphthongs, /ɪə eə ʊə əʊ aʊ eɪ əʊ/, with their first element in the general region of [ɪ e a ə ʊ] and the second element in the area of [ɪ ʊ ə]. Considering the direction of the glide, the RP set is classified into three centring diphthongs, /ɪə eə ʊə/, in which the articulation moves from the periphery of the CVS towards a mid-central neutral vowel [ə], and five closing diphthongs, in which the articulation either moves towards the near front, close, spread vowel [ɪ], /aɪ eɪ ɔɪ/, or towards the near close back rounded vowel [ʊ], /əʊ aʊ/. Turning to distance, RP has only three wide diphthongs, /aɪ ɔɪ aʊ/, where the distance between the two elements of the glide is bigger than in narrow diphthongs, which is the case of all the other RP diphthongs, whether centring, /ɪə eə ʊə/, or closing, /eɪ əʊ/. If we move on to prominence, all RP diphthongs are falling, that is, most of the length and stress associated with the glide is concentrated on the first element, the second element being only slightly sounded. Note also that /eɪ aɪ ɪə eə/ have unrounded lips throughout their articulation, while in the diphthongs /ɔɪ ʊə/ they change from rounded to unrounded, and in /əʊ aʊ/ they move from unrounded to rounded. In addition, the following generalisations may be applied to all the RP diphthongs:

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

122

Vowels and Vowel Glides

(1) With the exception of /ɔɪ/, they principally derive from former (typically long) pure vowels. (2) They are particularly susceptible to variation regionally and socially in both elements. The exact end-point of the glide depends on context, and can vary from speaker to speaker. (3) No diphthong occurs before /ŋ/, except where word-final /n/ is assimilated to /ŋ/ in connected speech, e.g. own [əʊŋ] in own car. (4) Before dark l [ɫ] diphthongs tend to be reduced to their corresponding first element (a normally long monophthong) plus a slight glide to [ə] or [ʊ], as in the case of e.g. failed [feəɫd], male [meəɫ], pail [peəɫ], sails [seəɫz]; aisle [aəɫ], mild [maəɫd], mile [maəɫ], piles [paəɫz]; cowl (n) [kaʊɫ], foul [faʊɫ], owl [aʊɫ]; bolt [bəʊɫt], hole [həʊɫ], moult [məʊɫt], roll [rəʊɫ], poles [pəʊɫz]; boils [bɔəɫz], coil [kɔəɫ], soil [sɔəɫ]; real [rɪəɫ]; cruel [ˈkruːəɫ] (see also fn 58 in § 3.3.3 on levelling or smoothing). (5) Note that, though very widespread, not all the speakers of British English use the diphthong [ʊə]. Instead, some use the long vowel [ɔː], or other diphthongs [ɔə]. Besides falling diphthongs, English also has rising glides, which arise when the first element (transcribed in LPD [j] or [w]) becomes the less prominent segment of the glide and is compressed with the following vowel into a single syllable. This occurs in the second syllable of such words as lenient and influence when they are pronounced fast: [ˈlɪːnjənt], [ˈɪnflwəns]. Faster or compressed pronunciations are more usual in frequently‑used words, in fast or casual speech, and if the word has already been used in the discourse. Alternatively, in slower pronunciations, especially those involving centring glides in which the schwa element represents a morphological ending, diphthongs tend to be realised as hiatuses with the two elements of the glide being pronounced over two different syllables and the first elements of /ɪə/ and /ʊə/ having a tense [i] and [u] realisation respectively. Accordingly, a slow pronunciation of lenient and influence would be [ˈlɪːni.ənt] and [ˈɪnflu.əns]. Another point worth of mention is that of hiatuses involving two instances of two similar vowel phonemes, which in English are neither fused or reduced. In such cases, foreign learners should be aware to pronounce two consecutive vowels, one tense and the other lax, as in, for instance, earliest /ˈɜːliɪst/ and employee /ɪmˈplɔɪiː/, rather than just one vowel (*[ˈɜːlɪst]) or a palatal semiconsonant plus a vowel (*[ɪmˈplɔjiː]). Lastly, it should be kept it mind that English tends to avoid the repetition of two identical vowels both within and between words. A good illustration of this is provided by the definite and the indefinite articles: to reinforce the hiatus that should exist between word boundaries, the is pronounced [ði], rather than [ðə],

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

123

if followed by vowels (the annoyance [ði əˈnɔɪəns], the apple [ði ˈapl]); whereas, in the same context, the indefinite article a [ə]/[eɪ] requires the presence of an alveolar nasal to demarcate word boundaries, and thereby becomes an [ən]/ [æn] (an annoyance [ən/æn əˈnɔɪəns], an apple [ən/æn ˈapl]). Turning to Spanish, it has fourteen diphthongs, /ai ̯ ei ̯ oi ̯ au̯ eu̯ ou̯ wa we wo wi ja je jo ju/, which are classified according to the closing-opening and falling-rising oppositions. Six are falling-closing /ai ̯ ei ̯ oi ̯, au̯ eu̯ ou̯], with the pattern V + semi-vowel, and the other eight are rising [ja je ju jo wa we wi wo], with the pattern semi-consonant + V, which are also opening except for [ju wi] (Quilis 1981; Quilis and Fernández 1982; Martínez Celdrán 1986). In addition, while English diphthongs have the phonological status of a long vowel and can be described as indivisible diphthongal nuclei that cannot be split by any phonological rule, Spanish diphthongs can be characterised as complex syllable nuclei consisting a sequence of a vowel + semivowel (falling-closing) or a semiconsonant + vowel (rising) that can be divided by a word-internal boundary (e.g. rey ‘king’ [rei ̯] → reyes ‘kings’ [re.ʝes]). Besides, in English [j w] are classified as two semi-consonant phonemes which can cluster with other consonants in onset clusters (e.g. [tw] in twenty /ˈtwenti/) and do not form diphthongs when followed by vowels (e.g. [jiː] yield /jiːld/, [wʊ] wood /wʊd/); in Spanish, by contrast, [j w] together with [i ̯] and [u̯ ] are generally regarded as allophonic realisations of /i/ and /u/ respectively and may create a diphthong or a triphthongs if followed by other vowels (e.g. radio ‘radio’ [ˈraðjo], cuadro [ˈkwaðɾo], buey ‘ox’ [bu̯ei ̯]) (Alarcos LLorach 1961 [1983]; Navarro Tomás 1966 [1946], 1991 [1918]; Quilis 1981; Martínez Celdrán 1986; Calvo Shadid 2008). Moreover, in Spanish diphthongs can also be pronounced as a result of resyllabification across words (e.g. jugó y comió ‘he played and ate’, [xu.ɣoi ̯.ko.mio]), or they may even disappear when the accent shifts (e.g. fiarás ‘(you) will trust’ vs. fías ‘(you) trust’). Nevertheless, despite that it may be tempting to believe that the Spanish sequences (día ‘day’, había ‘there was’), (tarea ‘task’, pelea ‘fight’) and (púa ‘spike’, ‘pick’, grúa ‘crane’) come close to the English vowel glides [ɪə], [eə] and [ʊə], respectively, they are in fact hiatuses, rather than diphthongs (Navarro Tomás 1991; Hualde 2005; Hualde and Prieto 2002).46 Triphthongs also occur in Spanish. They consist of two close vowels with a more open one in the middle and have the pattern semi-consonant + V + semi-vowel. They generally appear in verbs and result from the addition of inflectional endings (averigüéis [aβeɾiˈγwei ̯s] ‘you find out’ (pl pres subj), acariciáis [akaɾiˈθjai ̯s] ‘you caress’ (pl pres ind)) (Quilis and Fernández 1982). 46 The diphthong-hiatus distinction is based on changes in formant trajectories (especially of F2) and duration: hiatuses show a greater degree of curvature of the F2 trajectory and also longer duration than diphthongs, and as a consequence the vowels in a hiatus sequence, unlike those of diphthongs, belong to different syllables (Aguilar 1999: 72).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

124

Vowels and Vowel Glides

Figures 42 and 43 display starting points and directions for diphthongal vowel glides in RP and PSp, respectively. In what follows RP diphthongs and triphthongs are described along the same subsections that were considered for the description of vowels in Section 3.2. Note, however, that only RP closing diphthongs (Section 3.3.1) are contrasted with their PSp falling counterparts because centring diphthongs and triphthongs have no correlates in Spanish, as already advanced in Table 4 (Section 1.5).47 It will be shown that four RP diphthongs /aʊ eɪ aɪ ɔɪ/ seem to present no difficulties for SSLE as they tend to be assimilated with their Spanish counterparts /au̯ ai ̯ ei ̯ oi ̯/, whereas RP /əʊ/ is identified either with Spanish /ou/ or /o/ (Finch and Ortiz Lira 1982: 42; Gallardo del Puerto 2005: 186). Nevertheless, mispronunciations or communication flaws may arise broadly for three main reasons (Monroy Casas 2012, 1980; Estebas Vilaplana 2009; Swan and Smith 2001): (1) because more emphasis is laid on the second rather than on the first vowel of diphthongs due to the influence of Spanish; (2) because the last element of centring diphthongs (/ɪə ʊə eə/) is replaced by a flap [ɾ], probably due to the fact that most of these diphthongs are spelt with an , even though RP is a non-rhotic accent, in addition to the aforementioned difficulties of Spanish speakers in pronouncing the English schwa /ə/; (3) or because vowel glides, particularly closing diphthongs (/eɪ ɔɪ aɪ əʊ aʊ/), are monophthongised as a result of smoothing, despite the fact that, as already noted, this phenomenon affects the pronunciation of most glides amongst native speakers of English, especially in rapid speech.

Figure 42: RP diphthongs: (a) closing (b) centring

47 It could be argued that RP /ʊə/ is close to the PSp diphthong [wa] (e.g. suave ‘soft’, guante ‘glove’). However, this comparison is disregarded here because of their dissimilar nature: the Spanish diphthong is rising and the first element is less prominent, and the second element is more prominent, more open and fronted than in English.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

125

Figure 43: PSp diphthongs: (a) falling (b) rising

3.3.1 Closing diphthongs This section focuses on RP five closing diphthongs. Section 3.3.1.1 describes [ɪ]diphthongs, /eɪ aɪ ɔɪ/, displayed in Figure 44 along with their PSp counterparts, in which the articulation moves towards the near front, close, spread vowel [ɪ] or [i]. Section 3.3.1.2, on the other hand, characterises [ʊ]-diphthongs, [əʊ aʊ], presented in Figure 45 in comparison with their PSp counterparts, where there is a vowel glide towards the near close back rounded vowel [ʊ] or [u]. Note that /eɪ, aɪ/ have unrounded lips throughout their articulation, while in /ɔɪ ʊə/ they change from rounded to unrounded and in /əʊ aʊ/ they move from unrounded to rounded.

3.3.1.1 RP /eɪ aɪ ɔɪ/ vs. PSp /ei ̯ ai ̯ oi ̯/ IPA Symbols eɪ Lower-case E (or Latin small letter E) plus Small capital I (or Latin letter small capital I) combination. aɪ Lower-case A (or Latin Small letter A) plus Small capital I (or Latin letter small capital I) combination. ɔɪ Latin small letter open O (or Latin small letter E) plus Small capital I (or Latin letter small capital I) combination. ei ̯ Lower-case E (or Latin small letter E) plus plus reduced Lower case I (or Latin small letter I) combination.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

126 ai ̯ oi ̯

Vowels and Vowel Glides

Lower-case A (or Latin Small letter A) plus plus reduced Lower case I (or Latin small letter I) combination. Latin small letter O plus reduced Lower-case I (or Latin small letter I) combination.

Identification [eɪ] Near-close front, falling, narrow, closing diphthong (unrounded). [aɪ] Near-close front, falling, wide, closing diphthong (unrounded). [ɔɪ] Near-close front, falling, wide, closing diphthong (rounded to unrounded). Position in the CVS

Figure 44: [ɪ/i] diphthongs in RP (a) and PSp (b)

Description For the articulation of RP /eɪ/, the glide begins from slightly below the half-close front position [e̞] and moves in the direction of RP /ɪ/, there being a slight closing movement of the lower jaw. The lips are spread throughout. RP /aɪ/, on the other hand, begins at a point slightly behind the front open position [ä], approaching RP /ɑ/, and moves in the direction of RP /ɪ/ nearer to a centralised half-close position [ё̞]. The closing movement of the lower jaw is obvious and the lips change from a neutral to a loosely spread position. RP /ɔɪ/, in turn, starts with the centre of the tongue raised between halfopen and half-close, and then it moves in the direction of RP /ɪ/ nearer to a centralised half-close position [ё̞]. There is a slight closing movement of the lower jaw, and the lips are neutral for the first element but have a tendency to round on the second segment. The starting point may have a tongue position similar to RP /ɒ/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

127

See the animations and videos under RP /eɪ/, /aɪ/ and /ɔɪ/ (compared with Spanish /ei ̯ ai ̯ oi ̯/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 128–133, 134–138; Gimson 1994: 120–123, 124–127; Roach 2005: 21–23; Collins and Mees 2009: 102–104; Cruttenden 2014: 140–149). Spellings Table 14: The spellings of RP /eɪ aɪ ɔɪ/ /eɪ/ (in stressed syllables)

/aɪ/ (in stressed syllables)

/ɔɪ/ (in stressed syllables)

48 April /ˈeɪprəl/ Cambridge /ˈkeɪmbrɪʤ/ chaos /ˈkeɪɒs/

49 idle /ˈaɪdl ̩/, life /laɪf/

boil /bɔɪl/, toilet /ˈtɔɪlɪt/

50 jail /ʤeɪl/, rain /reɪn/ crayon /ˈkreɪən/

51 lie /laɪ/ bye /baɪ/, dry /draɪ/ asylum /əˈsaɪləm/ scythe /saɪð/

joy /ˈʤɔɪ/, oysters /ˈɔɪstəz/

52 grey /ɡreɪ/ obey /əˈbeɪ/

either /ˈaɪðə/ neither /ˈnaɪðə/ eye /aɪ/

53 great /ˈɡreɪt/ steak /steɪk/

+ silent 54 bright /braɪt/ high /haɪ/

+ silent sleigh /sleɪ/, eight /eɪt/

buy /baɪ/, guy /ɡaɪ/

in 55 graduate /ˈɡræʤueɪt/ approximate /əˈprɒksɪmeɪt/ estimate /ˈestimeɪt/

(Lat, pl) alumni /əˈlʌmnaɪ/ termini /ˈtɜːmɪnaɪ/ radii /ˈreɪdɪaɪ/

French loanwords

debut /ˈdeɪbjuː/ café /ˈkæfeɪ/ crepe /kreɪp/

matinee /ˈmætɪneɪ/ toupee /ˈtuːpeɪ/

ballet /ˈbæleɪ/ chalet /ˈʃæleɪ/

dossier /ˈdɒsɪeɪ/ foyer /ˈfɔɪeɪ/

German names and loanwords Kreutzer /ˈkrɔɪtsə/ Plattdeutsch /ˈplætdɔɪʧ/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

128

Vowels and Vowel Glides

Table 14: (Continued) /eɪ/ (in stressed syllables)

/aɪ/ (in stressed syllables)

/ɔɪ/ (in stressed syllables)

Rare cases Gaelic /ˈɡeɪlɪk/ gaol /ʤeɪl/ gauge /ɡeɪʤ/

Rare cases maestro /ˈmaɪstrəʊ/ aisle /aɪl/ aye /aɪ/ guise /ɡaɪz/

Rare cases buoy /bɔɪ/

Exceptions quay /kɪː/ said /sed/ says /sez/

Exceptions bewilder /bɪˈwɪlda/ children /ˈtʃɪldrən/ wilderness /ˈwɪldənɪs/ wind /wɪnd/ (n) (vs. wind /waɪnd/ (vb))

Exceptions choir /kwaɪə/

Regional and social variants The [ɪ] element of the vowel glides /eɪ, aɪ, ɔɪ/ tends to be absorbed into an [ə] or [ʊ] glide before dark [ɫ], as in sail [seəɫ]), tile [taːəɫ]) and boil [bɔːəɫ]. In the case of /eɪ/, older speakers may have a closer starting point nearer to CV2 [2], whereas Refined RP speakers prefer a more open starting point nearer to CV3 [ɛ]. In addition, as a result of the “Southern diphthong shift”, typical of London, Cockney, Birmingham, the South and Midlands of England, [eɪ] may be pronounced as [aɪ], [aɪ] as [ɑɪ] or [əɪ], and [ɔɪ] as [oi]. Besides, in Refined RP and Cockney (as well as in the so‑called mid‑Atlantic pronunciation) a very back starting point for [aɪ] is most common, [ɑ̈ɪ], which may sometimes involve the elimination of the glide, leaving a long monophthong [ɑɑː] (e.g. pop singers usually reduce I and my to [ɑː] and [mɑː]), while in a wide area of South-west 48 Note the different pronunciations and meaning of bass /beɪs/ (‘instrument’) and /bæs/ (‘fish’). 49 /aɪ/ is the NP of when it is followed by a simple C + silent . This is also the NP of in usually word-final (e.g. find /faɪnd/, blind /blaɪnd/, hind /haɪnd/), final (e.g. child /ʧaɪld/, wild /waɪld/, mild /maɪld/), and when followed by a silent , usually in wordfinal position (e.g. sign /saɪn/, malign /məˈlaɪn/, design /diˈzaɪn/). Note that if is not silent, then the pronunciation is /ɪ/ (e.g. signal /ˈsɪɡnl/, signature /ˈsɪɡnəʧə/). 50 If the sequences are followed by , their NP is /eə/. 51 are pronounced [aɪ] usually in word-final position, but also in some medial cases (e.g. cycle /ˈsaɪk(ə)l/, asylum /əˈsaɪləm/). 52 The spelling can also be pronounced /iː/ (e.g. conceive /kənˈsiːv/, perceive /pəˈsiːv/). 53 The NP of is /iː/ (e.g. jeans /dʒiːnz/, Jean / /dʒiːn/ (female name) vs. /dʒɑ:n/ (male name). 54 are pronounced [aɪ] usually in word-final position and/or before . 55 Beware that in the suffix is pronounced /eɪ/ in verbs, but in nouns and adjectives its pronunciation is /ət/ (e.g. /'ɡrædʒuət/, /əˈprɒksɪmət/, /ˈestimət/).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

129

and South-central England prefer [əɪ] pronunciations, with unrounded, raised, and centralised starting points. This is also the quality of refined RP pronunciations of /ɔɪ/, [əɪ] or [ɑɪ], which are characterised by unrounded, raised, and centralised starting points. Comparison with Spanish and advice It can be said that the three English [ɪ]-diphthongs, /eɪ/ (lay /leɪ/), /aɪ/ (high /haɪ/) and /ɔɪ/ (boy /bɔɪ/), approximate in quality to the Spanish ones, /ei ̯/ (ley ‘law’ [lei]), /ai ̯/ (hay ‘there is’ [ai ̯]) and /oi ̯/ (voy ‘(I) go’ [boi ̯]). RP /aɪ aʊ/, in particular, start in an area that is quite similar to that of Spanish /a/ and the starting points of /ai ̯ au̯/. RP /eɪ ɔɪ/, however, start in a lower area of the CVS than their Spanish counterparts /ei ̯ oi ̯/. But the main difference between the two languages resides in the second element: in Spanish it has a clear [i] quality, whereas in English the [ɪ] sound is more relaxed and open, between Spanish [i] and [e]. However, no misunderstandings will occur if a Spanish [i] type of vowel is uttered as the end point of these three diphthongs (Monroy Casas 1980, 2012; Gallardo del Puerto 2005: 184–186; Estebas Vilaplana 2009). Further practice AI 3.13. The spellings of RP /eɪ/. Further practice chaos /ˈkeɪɒs/ convey /kənˈveɪ/ explain /ɪkˈspleɪn/

plate /pleɪt/ played /pleɪd/ railway /ˈreɪlweɪ/

sleighing /ˈsleɪɪŋ/ they /ˈðeɪ/

The mayor says he’s afraid they elaborate a deliberate plan to raise riot in Baker street. ||ðə ˈmeə ˈsez hiz əˈfreɪd ðeɪ ɪˈlæbəreɪt ə dɪˈlɪbərət ˈplæn tə ˈreɪz ˈraɪət ɪn ˈbeɪkə ˈstriːt|| I don’t like this mayonnaise. May I change it? ||ˈaɪ dəʊnt ˈlaɪk ðɪs ˌmeɪəˈneɪz || ˈmeɪ aɪ ˈʧeɪnʤ ɪt || Today, at a quarter to eight, I’m going away to Spain by plane. ||təˈdeɪ | ət ə ˈkwɔːtə tu ˈeɪt | aɪm ˌɡəʊɪŋ əˈweɪ tə ˈspeɪn baɪ ˈpleɪn|| That’s a bit dangerous at your age. ||ˈðæts ə bɪt ˈdeɪnʤərəs ət jər ˈeɪʤ|| Six slimy snails sailed silently. ||ˈsɪks ˌslaɪmi ˈsneɪlz ˈseɪld ˈsaɪləntli|| On a lazy laser raiser lies a laser ray eraser. ||ɒn ə ˈleɪzi leɪzə ˈreɪzə ˈlaɪz ə ˈleɪzə reɪ ɪˈreɪzə||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

130

Vowels and Vowel Glides

AI 3.14. The spellings of RP /aɪ/. Further practice I /aɪ/ eyes /aɪz/ height /haɪt/

ice /aɪs/ island /ˈaɪlənd/ pie /ˈpaɪ/

tight /taɪt/ tried /traɪd/ typewriter /ˈtaɪpraɪtə/

Violet likes rock-climbing, horse-riding, and ice-skating. ||ˈvaɪələt ˈlaɪks ˈrɒkˌklaɪmɪŋ | ˈhɔːsˌraɪdɪŋ | ənd ˈaɪsˌskeɪtɪŋ|| Mike is going bicycle-riding with Miles next Friday. ||ˈmaɪk s ˈɡəʊɪŋ ˈbaɪsɪkl ̩ˌraɪdɪŋ wɪð ˈmaɪlz ˌnekst ˈfraɪdi|| I’ll ask the librarian if there are intriguing sci-fi and psycho movies on DVD. ||aɪl ˈɑːsk ðə laɪˈbreərɪən ɪf ðər ər ɪnˈtriːɡɪŋ ˈsaɪˌfaɪ ənd ˌsaɪkəʊ ˈmuːvɪz ɒn ˌdiviˈdiː|| Birdie birdie in the sky laid a turdie in my eye. ||ˈbɜːdi ˈbɜːdi ɪn ðə ˈskaɪ ˈleɪd ə ˈtɜːdi ɪn maɪ ˈaɪ|| If cows could fly, I’d have a cow pie in my eye. ||ɪf ˈkaʊz kəd ˈflaɪ | aɪd ˈhæv ə ˌkaʊ ˈpaɪ ɪn maɪ ˈaɪ||

AI 3.15. The spellings of RP /ɔɪ/. Further practice destroyed /dɪˈstrɔɪd/ flamboyant /flæmˈbɔɪənt/ Freud /frɔɪd/

noise /nɔɪz/ oyster /ˈɔɪstə/ tabloid /ˈtæblɔɪd/

toy /tɔɪ/ voice /vɔɪs/ voyage /ˈvɔɪɪʤ/

The engine of Joyce’s Rolls Royce makes an annoying noise, which is probably due to lack of oil. ||ði ˈenʤɪn əv ˈʤɔɪsəz ˌrəʊlz ˈrɔɪs ˈmeɪks ən əˈnɔɪɪŋ ˈnɔɪz | wɪʧ ɪz ˈprɒbəbli djuː tə ˈlæk əv ˈɔɪl|| Spoilt boys usually enjoy destroying noisy toys. ||ˌspɔɪlt ˈbɔɪz ˈjuːʒəli ɪnˈʤɔɪ dɪˈstrɔɪɪŋ ˌnɔɪzi ˈtɔɪz|| What noise annoys an oyster most? A noisy noise annoys an oyster most. ||ˈwɒt ˈnɔɪz əˈnɔɪz ən ˈɔɪstə ˈməʊst|| ə ˌnɔɪzi ˈnɔɪz əˈnɔɪz ən ˈɔɪstə ˈməʊst||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

131

3.3.1.2 RP /əʊ aʊ/ vs. PSp /eu̯ ou̯ au̯/

Figure 45: [ʊ/u] diphthongs in (a) RP and (b) PSp

IPA əʊ aʊ eu̯

Symbols Schwa and Lower-case Upsilon combination. Lower-case A and Lower-case Upsilon combination. Lower-case E (or Latin small letter E) and reduced Latin small letter U combination. ou̯ Latin small letter O and reduced Latin small letter U combination. au̯ Lower-case A (or Latin Small letter A) and reduced Latin small letter U combination. Identification [əʊ] Near-close back, falling, back, closing, narrow diphthong (unrounded to rounded). [aʊ] Near-close back, falling, back, closing, wide diphthong (unrounded to rounded). Description For the articulation of RP /əʊ/ the tongue moves from a centralised, between half-open and half-close position towards that of RP /ʊ/. There is a slight closing movement of the lower jaw, and the lips are neutral for the first element but tend to round on the second element. The starting point may have a tongue position similar to RP /ɜː/. By contrast, RP /aʊ/ begins at a point between the back and front open positions, slightly more fronted than the position for RP /ɑː/, and moves in the direction of RP /ʊ/ approaching the half-close level [ö]. The glide is much more extensive than that used for /əʊ/ and is symmetrically opposed to the front RP diphthong /aɪ/. The lips change from neutrally open to weakly rounded.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

132

Vowels and Vowel Glides

See the animations and videos under RP /əʊ/ and /aʊ/ (compared with Spanish /eu̯ ou̯ au̯/) in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 128–133, 134–138; Gimson 1994: 120–123, 124–127; Roach 2005: 21–23; Collins and Mees 2009: 102–104; Cruttenden 2014: 140–149). Spellings Table 15: The spellings of RP /əʊ aʊ/ /əʊ/ (in stressed syllables)

/aʊ/ (in stressed syllables)

cold /kəʊld/, won’t /wəʊnt/

council /ˈkaʊnsil/, loud /laʊd/

doe /dəʊ/, woe /wəʊ/

frown /fraʊn/, bow /baʊ/

/ + silent , ow shoulder /ˈʃəʊldə/, poultry /ˈpəʊltri/ though /ðəʊ/, sparrow /ˈspærəʊ/

+ silent drought /draʊt/, slough /slaʊ/

load /ləʊd/, oatmeal /ˈəʊtmiːl/ French loanwords

château /ˈʃætəʊ/, plateau /ˈplætəʊ/ bureau /ˈbjʊərəʊ/

mauve /məʊv/, au pair /əʊ peə/ gauche /ɡəʊʃ/

argot /ˈɑːɡəʊ/ , tarot /ˈtærəʊ/ Rare cases sew /səʊ/ brooch /brəʊʧ/ Pharaoh /ˈferəʊ/

Rare cases MacLeod /məˈklaʊd/

Regional and social variants English /əʊ/ involves a lot of variation. Conservative RP uses a rounded first element [öʊ], while Refined RP tends to produce an unrounded and advanced first element, between half-close and half-open, and centralised from front [ɛ̈ʊ]. Alternatively, an unrounded central monophthong can also be found [əː] (especially where [ɫ] follows). /əʊ/ is regularly kept in RP in unaccented syllables, but in colloquial speech and other “substandard” dialects it is generally reduced to

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

133

/ə/ (e.g. fellow [ˈfelə], obey [əˈbeɪ], phonetics [fəˈnetɪks], window [ˈwində]). This latter tendency results in the production of a number of homophones, that is, words with different spellings that have an identical pronunciation, in colloquial speech that are distinct in a more formal style: ferment – foment [fəˈment], hypertension – hypotension [haɪpəˈtenʃən] (see also Chapters 4 and 7). In addition, Refined RP, Cockney and Australian English have [æ̈ ʊ], even [æɣ̈], with fronting of the first element and unrounding spreading to the second element, whereas other accents (SSE), in much of the north of England and GA) have a relatively pure vowel around CV [o], which is close to the [oː] pronunciation that may also be heard in Cockney and is typical of London Regional RP. Moving on to /aʊ/, RP variants mainly involve the fronting ([a̘]) or retraction ([ɑ]) of the starting point, but it can also entail its raising ([æ], [ɛ], [ə], [ʌ], [uː]). Fronting is typical of popular regional forms of speech (the London region included), while in Cockney and Australian English, the first element is fronted and raised to the [æ] or [e] varieties. Alternatively, in Cockney /aʊ/ may also be monophthongised to long [aː]. Refined RP also prefers an [aːː] extra-long realisation of the first element, especially in those contexts in which the diphthong has its fully long form. Comparison with Spanish and advice The pronunciation of RP /əʊ/ may be problematic for Spanish learners because [ə] does not form part of the Spanish vocalic system. Spanish speakers tend to produce this diphthong with an initial [o], thereby resembling the pronunciation of the Spanish diphthong [ou̯ ] found in the word bou [bou̯ ] (‘type of fishing’). This is an acceptable pronunciation in GA, in which the normal pronunciation of this diphthong is [oʊ] instead of [əʊ]. But in RP such pronunciations have a strong foreign accent. Accordingly, in order to sound like RP speakers, Spanish learners should strive to start this diphthong with an [ə]-sound, with a slight opening of the mouth as if for mouth breathing, the articulators moving in the direction of a relaxed type of [u] in the end-point. The closest Spanish diphthong to RP [əʊ] is /eu̯ / (e.g. neutro [ˈneu̯ tɾo] ‘neutral’) but the first element of the RP glide is more central and the second more relaxed. RP /aʊ/, on the other hand, should have no problem for Spanish learners since it is similar to the Spanish diphthong [au̯ ] found in the word pausa [ˈpau̯ sa] (‘pause’), the only difference being that the starting point is more retracted and the end-point less relaxed in Spanish. However, the pronunciation of a Spanish [u] as the end-point of these two diphthongs should not cause intelligibility problems (Monroy Casas 1980, 2012; Gallardo del Puerto 2005; Estebas Vilaplana 2009).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

134

Vowels and Vowel Glides

Further practice AI 3.16. The spellings of RP /əʊ/. Further practice brooch /brəʊʧ/ foe /fəʊ/ grows /ɡrəʊz/

gross /ɡrəʊs/ neurosis /njʊəˈrəʊsɪs/ reproach /rɪˈprəʊʧ/

so /ˈsəʊ/ soldier /ˈsəʊlʤə/ though /ðəʊ/

Although you offered Joe Jones a soup bowl full of chocolate scones, he’s joking he’ll make a snowball and will throw it at your nose. ||ɔːlˈðəʊ ju ˈɒfəd ˈʤəʊ ˈʤəʊnz ə ˌsuːp ˈbəʊl ˈfʊl əv ˌʧɒklət ˈskɒnz | hiz ˈʤəʊkɪŋ hil ˈmeɪk ə ˈsnəʊbɑːl ənd wɪl ˈθrəʊ ɪt ət jə ˈnəʊz|| Hello, Joan, do you know that you can see the vocal folds with this microscope? ||həˈləʊ | ˈʤəʊn | də ju ˈnəʊ ðət ju kən ˈsiː ðə ˌvəʊkl ̩ ˈfəʊldz wɪð ðɪs ˈmaɪkrəskəʊp|| Oh! Moses supposes his toeses are roses, but Moses supposes erroneously. ||ˈəʊ | ˈməʊzɪz səˈpəʊzɪz ɪz ˈtəʊzɪz ə ˈrəʊzɪz | bət ˈməʊzɪz səˈpəʊzɪz ɪˈrəʊnɪəsli|| The two-toed tree toad tried to tread where the three-toed tree toad trod. ||ðə ˌtuːtəʊd triː ˈtəʊd ˈtraɪd tə ˈtred weə ðə ˌθriːtəʊd triː ˈtəʊd ˈtrɒd||

AI 3.17. The spellings of RP /aʊ/. Further practice a louse /ə laʊs/ allows /əˈlaʊz/ drought /ˈdraʊt/

how /haʊ/ mountain /ˈmaʊntɪn/ mouthe /ˈmaʊð/

pronounce /prəˈnaʊns/ renown /rɪˈnaʊn/ round-eyed /raʊndaɪd/

‘Turn the brown couch upside-down without doubts’, shouted Paul as loudly as he could. ||tɜːn ðə braʊn kaʊʧ ˈʌpsaɪd daʊn wɪðˈaʊt daʊts | ˈʃaʊtɪd pɔːl əz ˈlaʊdli əz hi kʊd|| The doughty clown raised his eye-brows when he was allowed to take the plough by the bough, and then browsed his foul task. ||ðə ˈdaʊti klaʊn reɪzd ɪz aɪ braʊz wen hi wəz əˈlaʊd tə teɪk ðə plaʊ baɪ ðə baʊ | ənd ðen braʊzd ɪz faʊl tɑːsk|| How much ground would a groundhog hog, if a groundhog could hog ground? ||ˈhaʊ ˈmʌʧ ɡraʊnd wʊd ə ˈɡraʊndˌhɑːɡ hɒɡ | ɪf ə ˈɡraʊndˌhɑːɡ kəd hɒɡ ɡraʊnd|| 3.3.2 Centring diphthongs: RP /ɪə eə ʊə/ This section concentrates on the three RP centring diphthongs, /ɪə/, /eə/, /ʊə/, in which the articulation moves from the periphery of the CVS towards a midcentral neutral vowel [ə]. The lips are unrounded throughout the articulation of /ɪə, eə/ and change from rounded to unrounded in /ʊə/ (see also e.g. Gimson

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

135

1984: 142–145; Gimson 1994: 131–134; Roach 2005: 22; Collins and Mees 2009: 104; Crutenden 2014: 153–156).

Figure 46: Centring diphthongs in RP

IPA ɪə eə ʊə

Symbols Small capital I (or Latin letter small capital I) and Schwa combination. Lower-case E and Schwa combination. Lower-case Upsilon and Schwa combination.

Identification [ɪə] Centring, falling, narrow (unrounded). [eə] Centring, falling, narrow (unrounded). [ʊə] Centring, falling, narrow (rounded to unrounded). Description The glide of RP /ɪə/ moves from a half-close and centralised front tongue position (as for RP /ɪ/) in the direction of the more open variety of /ə/ in word final positions, or to a mid type of /ə/ in non final contexts. The lips change from slightly spread to neutrally open. For RP /eə/, in turn, the glide begins from slightly above the half-open front position [ɛ] and moves in the direction of the more open variety of /ə/, especially word-finally. There is a very slight opening movement of the lower jaw and the lips are neutrally open throughout. The starting point is a bit closer than RP /æ/. For the articulation of RP /ʊə/ the tongue glides from a position similar to that used for /ʊ/ towards the more open type of /ə/ when occurring word-finally, and to a closer variety of /ə/ word medially. The lips change from weakly rounded to neutrally spread as the glide progresses.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

136

Vowels and Vowel Glides

See the animations and videos under RP /ɪə/, /eə/ and /ʊə/ in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 142–145; Gimson 1994: 131–134; Roach 2005: 22; Collins and Mees 2009: 104; Crutenden 2014: 153–156). Spellings Table 16: The spellings of RP /ɪə eə ʊə/ /ɪə/ (in stressed syllables)

/eə/ (in stressed syllables)

/ʊə/ (in stressed syllables)

+ / hero /ˈhɪərəʊ/, era /ˈɪərə/ mere /mɪə/, query /ˈkwɪəri/ experience /ɪkˈspɪərɪəns/

+ / + / Mary /ˈmeəri/, aerobic /eəˈrəʊbɪk/ care /keə/

/ + sure /ʃʊə/, cure /kjʊə/ jury /ˈʤʊəri/

beer /bɪə/ engineer /ˌenʤɪˈnɪə/

air /eə/, fair /feə/

boor /bʊə/ spoor /spʊə/

year /jɜː/, tear (n) /ˈtɪə/ beard /bɪəd/

pear /peə/, tear (v) /ˈteə/ wear /weə/

56 euro /ˈjʊərəʊ/ Europe /ˈjʊərəp/

chandelier /ˌʃændəˈlɪə/ weird /wɪəd/ emir /eˈmɪə/ Algeciras /ˌælʒɪˈsɪərəs/

heir /eə/, their /ðeə/

there /ðeə/, where /weə/ French loanwords

souvenir /ˌsuːvəˈnɪə/ Classical languages loanwords

theological /ˌθiəˈlɒʤɪkl ̩/

spontaneous /spɒnˈteɪniəs/

petroleum /pɪˈtrəʊliəm/

union /ˈjuːniən/ previous /ˈpriːviəs/

stadium /ˈsteɪdiəm/ delirium /dɪˈlɪriəm/

French loanwords

premiere /ˈpremieə/ Spanish loanwords

sombrero /sɒmˈbreərəʊ/

French loanwords

tour /tʊə/, dour /dʊə/ bourgeois /ˈbʊəʒwɑː/ courgette /kʊəˈʒet/ gourmet /ˈɡʊəmeɪ/ gourd /ɡʊəd/

liqueur /lɪˈkjʊə/

56 Here we can see that the pronunciations [jʊə] – [ju:ə] with /j/ preceding /ʊə/ correspond to the letter followed and/or preceded by the previous spellings.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

137

Table 16: (Continued) /ɪə/ (in stressed syllables)

Exceptions there /weə/ where /ðeə/

/eə/ (in stressed syllables)

/ʊə/ (in stressed syllables)

Rare cases mayor /meə/ prayer /preə/ Verdi /ˈveədi/

Rare cases

sewer 'drain' /ˈsʊə/

Exceptions are /ɑː/ weir /wɪə/ weird /wɪəd/

Exceptions truant /ˈtruːənt/

Note that in the case of /ʊə/, many of the words ending in , may also be pronounced [ɔə] or [ɔː] (e.g. poor [pʊə], [pɔː] (like paw (n, v)); sure [ ʃʊə], [ ʃɔə], [ ʃɔː] (like shore, Shaw)). Regional and social variants In unaccented syllables where the schwa element usually representing a suffix with morphemic status, the starting point of /ɪə/ and /ʊə/ may be pronounced as a closer and more tense vowel [iə] (e.g. easier, serious) and [uə] or [juːə] (e.g. influence, valuable, vacuum, eventual), which seems to suggest a hiatus of the vowels in sequence. Nowadays, these three diphthongs are increasingly monophthongised in RP. /ɪə/ tends to be pronounced [ɜː] or [ɑː] in Refined RP, especially in accented final syllables, which contrasts with the glide from a relatively close to an almost open position that may be heard in Cockney, sometimes with an intervening [j], i.e. [ijä]. /eə/ has a completely acceptable open monophthongal realisation [ɛ] in General RP, although Refined RP keeps the diphthong but with a more open starting point [æə̞], which contrasts with the closer starting point [e̝ə] of Cockney. With regard to /ʊə/, it is being gradually replaced by the long monophthong /ɔː/, although [ɔə] pronunciations may also be heard. Accordingly, the pronunciations of such words as Shaw, sure, shore may still be kept distinct by some speakers [ʃɔː], [ʃʊə], [ʃɔə], or they may be levelled to [ ʃɔː] by others; likewise, you’re may be realised as [jʊə] (most frequently) or as [jɔː] (identical with your). Nevertheless, commonly used monosyllabic words (e.g. dour /dʊə/, gourd /gʊəd/), the [uːə]/[uə]/[ʊə] derived from /uː/ plus /ə/ (e.g. pursuer /pəˈsuːə/), and words with a preceding [j] (e.g. bureau [ˈbjʊərəʊ], cure [kjʊə], curious [ˈkjʊərɪəs], endure [ɪnˈdjʊə], puerile [ˈpjʊəraɪl], secure [sɪˈkjʊə]) are generally not subject to lowerings or monophthongisations.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

138

Vowels and Vowel Glides

Comparison with Spanish and advice Spanish has no centring diphthongs,57 but approximate realisations of RP /ɪə/, /eə/ and /ʊə/ may be found in the hiatus sequence [ía] (e.g. día ‘day’, vía ‘way’, mía ‘mine’ (fem. sing.)), [eə] (e.g. pelea ‘fight’, gragea ‘pill’) and [úa] (e.g. púa ‘spike’), the starting point being closer, and the end-point, more prominent and more open and advanced in Spanish so that the glide is longer in Spanish than in English (Finch and Ortiz Lira 1982: 42; Gallardo del Puerto 2005: 186). Nevertheless, if the aim is to produce centring diphthongs with an RP accent, Spanish learners should pay particular attention to two aspects of pronunciation (Monroy Casas 1980, 2012; Estebas Vilaplana 2009): (1) to avoid the production of final flap [ɾ] because in most cases centring diphthongs are spelt with an and RP is a non-rhotic accent; and (2) to pronounce an [ə] vowel as the end-point of these diphthongs, which is produced with almost no movements of the articulators apart from the narrow opening of the mouth, instead of the vowels [e] or [a], which would be the cause of a strong foreign accent. Further practice AI 3.18. The spellings of RP /ɪə/. Further practice dear /dɪə/ fakir /ˈfeɪkɪə/ fears /fɪəz/

fierce /fɪəs/ material /məˈtɪərɪəl/ museum /mjuːˈzɪəm/

period /ˈpɪərɪəd/ sincere /sɪnˈsɪə/ zero /ˈzɪərəʊ/

‘What about having a beer and some cereal here, in the same place as last year?’ said Mr Lear to the Austrian, white bearded mountaineer. ||ˈwɒt əˈbaʊt ˈhævɪŋ ə ˈbɪər ənd səm ˈsɪərɪəl ˈhɪə | ɪn ðə seɪm ˈpleɪs əz ˌlɑːst ˈjɪə| ˈsed ˈmɪstə ˈlɪə tə ði ˌɒstriən | ˌwaɪtbɪədɪd ˌmaʊntɪˈnɪə|| Good idea, the atmosphere is ideal! ||ˈɡʊd aɪˈdɪə | ði ˈætməsfɪə z aɪˈdɪəl|| Here come the beers, cheers! ||ˈhɪə ˈkʌm ðə ˈbɪəz| ˈʧɪəz|| Real weird rear wheels! ||ˌrɪəl wɪəd ˌrɪə ˈwiːlz|| 57 English rhotic accents such as Scottish English or GA have no centring diphthongos either. Instead, they have the first element of the glide followed by /r/ in those words which have an in the spelling before consonants and before a pause: /ɪə/ is pronounced /ɪ/ or /ɪ/ + /r/; /eə/ results in /e/ or /eɪ/ + /r/ combinations; and /ʊə/ is realized as /uː/ or /ʊ/ + /r/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

139

Deer, deer, oh dear, your career as a deer is over here. ||ˈdɪə | ˈdɪə | əʊ ˈdɪə | jə kəˈrɪər əz ə ˈdɪə z ˈəʊvə hɪə|| Near an ear, a nearer ear, a nearly eerie ear. ||ˈnɪər ən ˈɪə | ə ˌnɪərər ˈɪə | ə ˈnɪəli ˌɪəri ˈɪə||

AI 3.19. The spellings of RP /eə/. Further practice aquaarium /əˈkweəriəm/ bear /beə/ boleros /bəˈleərəʊz/

mayor /meə/ millionaire /ˌmɪlɪəˈneə/ prayers (thing) /preəz/

scarce /skeəs/ scares /skeəz/ where /weə/

Claire has carefully looked for her square hairbrushes everywhere, upstairs and downstairs, but they’re nowhere. ||ˈkleə həz ˈkeəfəli ˈlʊkt fə hə ˌskweə ˈheəbrʌʃɪz ˈevrɪweə | ˌʌpˈsteəz ən ˌdaʊnˈsteəz | bət ˈðeə ˈnəʊweə|| My parents found it weird that a rare canary, a hare and a mare were eating a pair of their pears on the pier. ||maɪ ˈpeərənts ˈfaʊnd ɪt ˈwɪəd ðət ə ˌreə kəˈneəri | ə ˈheər ənd ə ˈmeə wər ˈiːtɪŋ ə ˈpeər əv ðeə ˈpeəz ɒn ðə ˈpɪə|| Fuzzy Wuzzy was a bear, Fuzzy Wuzzy had no hair. ||ˈfʌzi ˈwʌzi wəz ə ˈbeə | ˈfʌzi ˈwʌzi həd nəʊ ˈheə||

AI 3.20. The spellings of RP /ʊə/. Further practice actually /ˈækʧuəli/ /ˈæktjuəli/ neurological /ˌnjʊərəˈlɒʤɪkl ̩/ ritual /ˈrɪʧʊəl/ endurance /ɪnˈdjʊərəns/ poor /pʊə/ spurious /ˈspjʊəriəs/ jewel /ˈʤu:əl/ puerile /ˈpjʊəraɪl/ tourney /ˈtʊəni/ During my visit to the aestheticist to check my eye-contour and my manicure it was pouring rain in the moors. ||ˈdjʊərɪŋ maɪ ˈvɪzɪt tə ðə iːsˈθetɪcɪst tə ˈʧek maɪ ˌaɪˈkɒntʊə ənd maɪ ˈmænɪkjʊər ɪt wəz ˈpɔːrɪŋ ˈreɪn ɪn ðə ˈmʊəz|| Curiously Duracell batteries seem to have more endurance. ||ˈkjʊərɪəsli ˈdjʊərəˌsel ˈbætriz ˈsiːm tə ˈhæv mɔːr ɪnˈdjʊərəns|| The furious soldier’s shoulder surely got hurt in the tour and probably will have no cure. ||ðə ˈfjʊərɪəs ˌsolʤərz ˈʃəʊldə ˈʃʊəli ɡɒt ˈhɜːt ɪn ðə ˈtʊər ənd ˈprɒbəbli wl ˈhæv nəʊ ˈkjʊə||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

140

Vowels and Vowel Glides

3.3.3 Diphthongs + [ə] Closing diphthongs, /eɪ aɪ ɔɪ əʊ aʊ/, may be followed by /ə/, either as an inseparable part of the word (e.g. rayon /reɪən/, fire [faɪə], Noah [ˈnəʊə], soya [ˈsɔɪə], hour [ˈaʊə],) or as a morpheme that is appended to a root (e.g. gayer [geiə], higher [haɪə], slower [sləʊə], employer /ɪmˈplɔɪə/), in which case the second element /ɪ, ʊ/ is less prominent than the first and last. Five vowel glides may be obtained in this way: [eɪə aɪə ɔɪə əʊə aʊə]. In all of them the tongue begins a glide in the direction of the second element, but then the direction changes again towards a final central position, as shown in Figure 47 below (see also e.g. Gimson 1984: 139–142; Gimson 1994: 128–130; Roach 2005: 24–26; Collins and Mees 2009: 102, 103; Crutenden 2014: 150–152).

Figure 47: Diphthongs + [ə] in RP

It should be noted, however, that in those cases in which these glides spread over two syllables or more (e.g. gayer, player, slower), it would be inaccurate to regard them as a triphthongs, as they rather behave as hiatuses. Description In the phonemic combination in which the diphthong /eɪ/ is followed by /ə/ (as in gayer [geɪə]), the diphthongal glide is abruptly interrupted, and then the tongue moves with an audible glide to a relatively open variety of /ə/. In turn, when the diphthong /aɪ/ is followed by /ə/ (as in fire [ˈfaɪə]), the tongue glides from a position just behind and above front open, the lips being neutral, towards

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

141

a centralised front position just above half close, the lips becoming gradually slightly spread; but before reaching the half close height, the tongue then glides to a central position just below half open, the lips returning to a neutral position at the end of the glide. The movement of the tongue is therefore extensive. In the case of [ɔɪə] (as in joyous [ˈʤɔɪəs]) and [əʊə] (as in slower [sləʊə]), the diphthongal glides ([ɔɪ], [əʊ]) are abruptly interrupted, and then the tongue moves with an audible glide to a relatively open variety of /ə/. In the case of [ɔɪə], the lips change from open rounded to neutral, and from neutral to round and then back to neutral in [əʊə]. In both the movement of the tongue is not extensive. Lastly, for the articulation of [aʊə] (as in power [ˈpaʊə]), the tongue glides from just above a centralised open position towards a centralised back position just above half close, but before reaching the half close height, the tongue then glides to a central position just below half open. The lips change from neutrally open to weakly rounded and back to neutral at the end of the glide. The movement of the tongue is therefore extensive. See the animations and videos under RP /eɪə/, /aɪə/, /ɔɪə/, /əʊə/ and /aʊə/ in the Sound Bank of the EPSS Multimedia Lab (see also e.g. Gimson 1984: 139–142; Gimson 1994: 128–130; Roach 2005: 24–26; Collins and Mees 2009: 102, 103; Crutenden 2014: 150–152). Spellings Table 17: The spellings of RP diphthongs + [ə] Diphthongs + [ə]

Spellings

/eɪə/

player /pleɪə/, layer /ˈleɪə/

/aɪə/

tired /ˈtaɪəd/ tyre /ˈtaɪə/, lyre /ˈlaɪə/ flier /ˈflaɪə/ via /ˈvaɪə/

/ɔɪə/

destroyer /dɪˈstrɔɪə/ employer /ɪmˈplɔɪə/ enjoyable /ɪnˈʤɔɪəbl ̩/

/əʊə/

mower /ˈməʊə/, slower /ˈsləʊə/

/aʊə/

shower /ˈʃaʊə/ flour /ˈflaʊə/, hour /ˈaʊə/

Rare cases choir /ˈkwaɪə/ virus /ˈvaɪərəs/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

142

Vowels and Vowel Glides

Regional and social variants Given that the second element of these vowel glides is very weak, there is a tendency in General RP and in Refined RP, especially if spoken rapidly, to drop this second element thereby reducing the glide to two elements (diphthong) or to one element (monophthong) so that such words as layer and lair may be pronounced as [lɛː] and [laː], respectively. This process is referred to as levelling or smoothing58 and it results in the following realisations: (1) [aɪə] → [aːə] as in buyer [ˈbaːə], choir [ˈkwaːə], fire [faːə], higher [ˈhaːə], liar [ˈlaːə], liable [ˈlaːəb(ə)l], lyre [laːə], society [səˈsaːəti], shyer [ˈʃaːə], tyre [taːə]. Diphthongal pronunciations are often reduced to the long monophthong [aː]. (2) [aʊə] → [ɑːə] as in coward [ˈkɑːəd], nowadays [ˈnɑːədeɪz], our [ˈɑːə], shower [ˈʃɑːə]. The qualitative difference distinguishing the first element of [aːə] and [ɑːə] is often levelled out to one central open vowel, resulting in the production of such homophones as sire-sour; shire-shower; tyre-tower. Further reductions may also occur so that [aʊə] may be pronounced as the long monophthong [ɑː]. In accents with a more extensive levelling (such as affected RP), this is also the pronunciation of the monophthongisations in (1) above, and as a result such [ɑː]-homophones may be heard: byre-buyerbower-bar [bɑː]; shire-shower-Shah [ ʃɑː]; tyre-tower-tar [tɑː]. (3) [eɪə] → [eːə] / [eə], as in conveyor [kənˈveːə], greyer [ɡreːə], layer [leːə], payer [ˈpeːə], while a monophthongisation to [ɛː] is a completely acceptable alternative in General RP. (4) [əʊə] → [ɜːə] → [ɜː],59 so that in General RP homophones may be produced such as mower-myrrh [ˈmɜː]; slower-slur [slɜː]. (5) [ɔɪə] → [ɔːə], as in buoyant [ˈbɔːənt], employer [ɪmˈplɔːə], enjoyable [ɪnˈʤɔːəb(ə)l], joyous [ˈʤɔːəs]. (6) Likewise, /eɪ, aɪ, ɔɪ/ tend to lose the [ɪ] element when followed by /iː/ or /ɪ/ both between and within words (e.g. playing, buy it). Comparison with Spanish and advice with PSp and advice 58 Smoothing, or the loss of the second part of a diphthong or triphthong, should be distinguished from the phenomenon of compression, or the squashing of the two syllables into one syllable. Both processes are optional, or stylistically determined. Hence, for instance, given the disyllabic starting point power [paʊ.ə], the word can be smoothed to disyllabic [pa.ə], and then the result can be compressed to [paə] or even further to a monophthong [paː]. Similarly, going [ɡəʊ.ɪŋ] can be smoothed to [ɡə.ɪŋ] and then compressed to [ɡəɪŋ]. 59 Note that when the first vowel of the triphthong is [ə] and it is levelled, the corresponding long monophthong is transcribed as [ɜˑ], because [ə] can only occur alone in unstressed syllables and therefore the sequence [əˑə] is reinterpreted as [ɜˑə].

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

143

As it happened with centring diphthongs, the main problem that RP triphthongs pose to Spanish speakers is grasping the correct pronunciation of the [ə] endpoint, for which special attention should be paid to: (1) avoiding the production of a final [r̄] when an is present in spelling, and (2) ending these glides with an [ə] vowel, by relaxing the articulators and opening the mouth slightly, instead of pronouncing the vowels [e] or [a] (in such words as tower or loyal), which would be the cause of a strong foreign accent (Monroy Casas 1980, 2012; Estebas Vilaplana 2009). Also, it is important that learners resist the phonotactic tendency of Spanish to palatalise [j] or velarise [w] the second element of the glide ([ɪ], [ʊ]) (as in fire *[ˈfajə], employer *[emˈplɔjə] and tower *[ˈtawə], which would amount to a foreign accent. Foreign learners should be aware of the reductions described above in order to be able to understand colloquial English, as well as to be able to reproduce the reductions that are normal among educated speakers. Nevertheless, as these reductions are often stigmatised as vulgarisms, extreme forms of levelling (monophthongisations) should be avoided, while unreduced pronunciations should be preferred in those cases in which the [ə] element corresponds to a suffix with a definite meaning, as in such words as higher /haɪə/ or plougher /plaʊə/. Further practice AI 3.21. The spellings of RP /eɪə/. Further practice conveyor /kənˈveɪə/ gaiaty /ˈgeɪəti/ greyer /ˈgreɪə/

layer /ˈleɪə/ mayonnaise /meɪəneɪz/ player /ˈpleɪə/

prayer (person) /ˈpreɪə/ stayer /ˈsteɪə/ rayon /reɪən/

The prayer and the player agreed that the Bible has multiple layers of meaning. ||ðə ˈpreɪər ənd ðə ˈpleɪər əˈɡriːd ðət ðə ˈbaɪbl ̩ həz ˈmʌltɪpl ̩ ˈleɪəz əv ˈmiːnɪŋ|| The payer in rayon is really a stayer but not a betrayer. ||ðə ˈpeɪər ɪn ˈreɪәn z ˈrɪəli ə ˈsteɪə bət nɒt ə bɪˈtreɪə||

AI 3.22. The spellings of RP /aɪə/. Further practice attire /əˈtaɪə/ buyer /ˈbaɪə/ flier /ˈflaɪə/

iron /ˈaɪən/ liar /ˈlaɪə/ psychiatrist /sɪˈkaɪətrɪst/

science /ˈsaɪəns/ tyre /ˈtaɪə/ wire /ˈwaɪə/

The clients found some iron wire in this tyre. ||ðə ˈklaɪənts faʊnd səm ˈaɪən ˈwaɪər ɪn ðɪs ˈtaɪə||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

144

Vowels and Vowel Glides

In Ireland some buyers hired a car to a liar without finding out if the price was higher or lower. ||ɪn ˈaɪələnd səm ˈbaɪəz ˈhaɪəd ə kɑː tu ə ˈlaɪə wɪðˈaʊt ˈfaɪndɪŋ aʊt ɪf ðə praɪs wəz ˈhaɪər ɔː ˈləʊə||

AI 3.23. The spellings of RP /ɔɪə/. Further practice buoyant /ˈbɔɪənt/ destroyer /dɪˈstrɔɪə/ employer /ɪmˈplɔɪə/

enjoyable /ɪnˈʤɔɪəbl ̩/ foyer /ˈfɔɪə/ joyous /ˈʤɔɪəs/

loyal /lɔɪəl/ soya /ˈsɔɪə/ Toya /ˈtɔɪə/

Roger met his joyous and flamboyant lawyer in the foyer. ||ˈrɒʤə met ɪz ˈʤɔɪəs ənd flæmˈbɔɪənt ˈlɔːjər ɪn ðə ˈfɔɪə|| Being an employer or an employee these days is not something enjoyable. ||ˈbiːɪŋ ən ɪmˈplɔɪər ɔːr ən ˌemplɔɪˈiː ðiːz deɪz ɪz nɒt ˈsʌmθɪŋ ɪnˈʤɔɪəbl ̩||

AI 3.24. The spellings of RP /əʊə/. Further practice blower /ˈbləʊə/ goer /ˈɡəʊə/ lower /ˈləʊə/

moa /ˈməʊə/ mower /ˈməʊə/ rower /ˈrəʊə/

shower /ʃəʊə/ ‘one who shows’ sower /ˈsəʊə/ sewer /ˈsəʊə/ ‘one who sews’

This rower is slower than the other. ||ðɪs ˈrəʊə z ˈsləʊə ðən ði ˈʌðə|| Meeting-goers and homeowners were offered blowers, lawn-mowers and shower-head nozzles at lower prices. ||ˈmiːtɪŋ ˈɡəʊərz ənd ˈhəʊməʊnəz wər ˈɒfəd ˈbləʊəz | lɔːn ˈməʊəz ənd ˈʃaʊə hed ˈnɒzl ̩z ət ˈləʊə ˈpraɪsɪz||

AI 3.25. The spellings of RP /aʊə/. Further practice coward /ˈkaʊəd/ flower /ˈflaʊə/ flour /ˈflaʊə/

our /ˈaʊə/ power /ˈpaʊə/ sour /ˈsaʊə/

towel /ˈtaʊəl/ tower /ˈtaʊə/ vowel /ˈvaʊəl/

The soldier with the beard that we saw in the bower market is not a hero, but a coward. ||ðə ˈsəʊlʤə wɪð ðə bɪəd ðət wi ˈsɔː ɪn ðə ˈbaʊə ˈmɑːkɪt s nɒt ə ˈhɪərəʊ | bət ə ˈkaʊəd||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

145

Don’t put the milk by the fire because it may go sour. ||dəʊnt ˈpʊt ðə mɪlk baɪ ðə ˈfaɪə bɪˈkɒz ɪt meɪ ɡəʊ ˈsaʊə|| The Shah always uses our flowery towel after the shower. ||ðə ʃɑː ˈɔːlweɪz ˈjuːsɪz ˈaʊə ˈflaʊəri ˈtaʊəl ˈɑːftə ðə ˈʃaʊə||

Further reading There exist many volumes that devote specific chapters or sections to the description of English vowels and vocalic glides. Again, we recommend Cruttenden (2014, Chapter 8), Ogden (2009, Chapter 5), Ladefoged (2001, Chapter 9; 2005, Chapter 15), Roach (2005, Chapters 2 and 3), Collins and Mees (2003, pp. 89– 126), Ladefoged and Maddieson (1996, Chapter 9) and Gimson (1984, Chapter 7), among others. A more advanced account of the acoustic features of vowels and vowel glides is offered by Hayward (2000, Chapter 6). Turning to Spanish, in addition to Cabrera Abreu and Vizcaíno Ortega’s (2009) acoustic-based analysis, classic descriptions of the acoustic features of Spanish vowels and vowel glides may be found in Martínez Celdrán (1998) and Quilis (1981), while more general descriptions are provided in Martínez Celdrán and Fernández Planas (2007), Navarro Tomás (1991 [1918], 1966 [1946]), Quilis and Fernández (1982), Quilis (1985, 1993), and Alarcos Llorach (1961 [1983]). Among the recommended texts that describe RP vowels specifically oriented to SSLE are, as already noted in the introduction, Monroy Casas (1980, 2012, Chapters1 and 2), Estebas Vilaplana (2009, Chapter 1), Mott (2011, Chapters 5 and 12) and Finch and Ortiz Lira (1982, Chapter 6), Lillo (2009, Chapter 2) and García Lecumberri and Maidment (2000: 8-16) providing transcription practice material. Finally, more detailed accounts concerning the acquisition of English vowels by speakers of Spanish are offered in Escudero and Chládková (2010), Gallardo del Puerto (2005), Escudero (2000), García Lecumberri and Cenoz Iragui (1997), and Dale and Poms (1985), to mention but a few.

Exercises 1. Passive transcription. Read aloud the following passage given in broad transcription, paying special attention to the vowel sounds. Then write it out in ordinary spelling. / nʌn əv ði ˈɔːɡənz kənˈfɔːmɪŋ ðə θriː ˈsɪstəmz ɪnˈvɒlvd ɪn spiːʧ prəˈdʌkʃn̩ hæv spiːʧ əz ðeə meɪn əˈrɪʤn̩əl ˈfʌŋkʃn̩ //fər ɪɡˈzɑːmpl ̩ / ðə lʌŋz ɑː fə ˈbriːðɪŋ / ðə ˈvəʊkl ̩ kɔːdz ɑː fə prɪˈventɪŋ ˈʧəʊkɪŋ / ðə tʌŋ ɪz fər ˈiːtɪŋ ənd ˈteɪstɪŋ / ðə nəʊz

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

146

Vowels and Vowel Glides

fə ˈbriːðɪŋ ən ˈsmelɪŋ / ən ˈsəʊ ɒn / bət ˈðeɪ həv biːn əˈdæptɪd tə prəˈdjuːs kə ˈmjuːnɪkətɪv ˈsaʊndz / əz ɪz ɪkˈspleɪnd ɪn ˈwɒt ˈfɒləʊz // 2. Explain the difference between the vowels of /fɪl/ – /fiːl/ and – /fʊl/ – /fuːl/. Why do we actually use two different symbols -i.e. /ɪ/ – /iː/ and /ʊ/ – /uː/to represent these vowel sounds, instead of just one, with or without the duration mark? 3. Group 1: /ɪ/ and /iː/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /ʧiːz/ 5. /bɪʧ/ 9. /liːv/ 13. /ðiːz/ 17. /pəˈliːs/

2. /ˈbɪskɪt/ 6. /wiːk/ 10. /kiːn/ 14. /ˈfɪzɪks/ 18. /ˈliːdə/

3. /piːʧ/ 7. /ʤiːnz/ 11. /ˈiːvnŋ/ 15. /mɪst/ 19. /kiː/

4. /stɪl/ 8. /ˈletɪs/ 12. /piːs/ 16. /gɪlt/ 20. /liːp/

4. Group 1: /ɪ/ and /iː/. Give a phonemic transcription of the words below and then classify them into two columns. Note that there is one word which does not fit in either column. women key reach meat sheet

wanted rich private receive busy

please Sunday tea people pleasure

village complete piece build pretty

minute tip city field sleep

5. Group 2: /e/, /ə/ and /ɜː/. Read aloud the following words and then write them out in ordinary spelling. 1. /ˈnekləs/ 4. /ˈbrekfəst/ 7. /bɜːθ/ 10. /ˈberɪ/ 13. /wɜːd/ 16. /ˈkʌlə/ 19. /bɜːd/

2. /ˈɒnə/ 5. /ɜːn/ 8. /ɜːθ/ 11. /gɜːl/ 14. /red/ 17. /ˈmʌðə/ 20. /ʧɜːʧ/

3. /ˈprefəs/ 6. /hɜːd/ 9. /wɜːk/ 12. /ˈweðə/ 15. /ˈʤɜːnɪ/ 18. /ɜːʤ/ 21. /pɜːl/

6. Group 2: /e/, /ə/ and /ɜː/. Give a phonemic transcription of the words below and then classify them into three columns. Note that there is one word which does not fit in any column. nurse river many journey church

best skirt bury search afraid

world head thirst person dead

stomach mother pearl them breath

turn friend work worm conservation

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

147

7. Group 3: /æ/, /ɑː/ and /ʌ/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /stʌf/ 4. /ˈbʌrə/ 7. /bæŋk/ 10. /ɑːnt/ 13. /sʌn/ 16. /wʌn/ 19. /drʌŋk/

2. /ˈfɑːðə / 5. /stɑːf/ 8. /lʌv/ 11. /dræŋk/ 14. /ˈbɑːskɪt/ 17. /ˈsæmən/ 20. /ˈkʌntri/

3. /ˈæplz/ 6. /blʌd/ 9. /rɪˈmɑːks/ 12. /nʌn/ 15. /ˈɑːnsə/ 18. /ænt/ 21. /lɑːfs/

8. Group 3: /æ/, /ɑː/ and /ʌ/. Give a phonemic transcription of the words below and then classify them into three columns. Note that there is one word which does not fit in any columns. duck month bad love grass

aunt hut bother past park

badge Monday uncle man country

hard laugh blood half glass

packet garden bath father heart

9. Group 4: /ɒ/ and /ɔː/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /pɔːt/ 4. /ˈwɔːtə/ 7. /tɔːk/ 10. /ˈɔːltə/ 13. /wɔːk/ 16. /bɔː/ 19. /pɔː/

2. /kɔːs/ 5. /sɔː/ 8. /kɔːt/ 11. /rɔː/ 14. /kɒst/ 17. /ɔː/ 20. /wɔːn/

3. /swɒn/ 6. /gɒn/ 9. /wɔːd/ 12. /wɒt/ 15. /sɒlt/ 18. /bɪˈkɒz/ 21. /jɔː/

10. Group 4: /ɒ/ and /ɔː/. Give a phonemic transcription of the words below and then classify them into two columns. Note that there is one word which does not fit in either column. horse porter cause because

doctor potter sausage fault

law salt wonder watch

forty caught bought daughter

wander war wash knowledge

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

148

Vowels and Vowel Glides

11. Group 5: /ʊ/ and /uː/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /pʊl/ 4. /gruːp/ 7. /fruːt/ 10. /gʊd/ 13. /wʊd/ 16. /bʊʃ/ 19. /ʧuːz/

2. /fluː/ 5. /bʊk/ 8. /ruːd/ 11. /luːk/ 14. /fʊt/ 17. /ʤuːs/ 20. /θruː/

3. /ruːt/ 6. /suːn/ 9. /wʊlf/ 12. /lʊk/ 15. /ʃuː/ 18. /fuːd/ 21. /kruːz/

12. Group 5: /ʊ/ and /uː/. Give a phonemic transcription of the words below and then classify them into two columns. Note that there is one word which does not fit in either column. butcher foot put pool moon

wolf pull soon clue good

full June move grew fool

butler true food sugar cushion

fruit rude cook woman Jew

13. Closing diphthongs: /eɪ/, /aɪ/, /ɔɪ/, /əʊ/ and /aʊ/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /reɪn/ 4. /həʊm/ 7. /əʊn/ 10. /rəʊd/ 13. /ðeɪ/ 16. /kraɪ/ 19. /heɪt/

2. /taɪm/ 5. /haʊs/ 8. /braʊn/ 11. /bɔɪ/ 14. /weɪt/ 17. /fraɪt/ 20. /aɪl/

3. /nɔɪz/ 6. /laʊd/ 9. /θrəʊ/ 12. /deɪ/ 15. /daɪ/ 18. /kɔɪn/ 21. /eɪt/

14. Closing diphthongs: /eɪ/, /aɪ/, /ɔɪ/, /əʊ/ and /aʊ/. Give a phonemic transcription of the words below and then classify them into columns. Note that there is one word which does not fit in any column. break toy sky plain though

favour mile point said town

plough road toe buy around

weight voice height know say

sign boy aid mouse brown

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

A comparison of English and Spanish vowel glides

149

15. Centring diphthongs: /ɪə/, /eə/ and /ʊə/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /dɪə/ 4. /weə/ 7. /ʃʊə/ 10. /fɪə/ 13. /mʊə/

2. /beə/ 5. /peə/ 8. /hɪə/ 11. /heə/ 14. /ˈhɪərəʊ/

3. /pʊə/ 6. /feə/ 9. /bɪəd/ 12. /reə/ 15. /tʊә/

16. Centring diphthongs: /ɪə/, /eə/ and /ʊə/. Give a phonemic transcription of the words below and then classify them into three columns. Note that there is one word which does not fit in any column. real mayor where parents

share jury stare clean

jewel cruel sure peer

idea year fierce hair

weird cure cheer tour

17. Diphthongs + [ə]: /aɪə/, /eɪə/, /ɔɪə/, /əʊə/ and /aʊə/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /ˈfraɪə/ 4. /ˈlɔɪəl/ 7. /ˈflaʊə/ 10. /ˈsləʊə/

2. /ˈpleɪə/ 5. /ˈtaʊə/ 8. /ˈrɔɪəl/ 11. /ˈdraɪə/

3. /ˈləʊə/ 6. /ˈleɪə/ 9. /ˈaɪən/ 12. /ˈgreɪə/

18. Diphthongs + [ə]: /aɪə/, /eɪə/, /ɔɪə/, /əʊə/ and /aʊə/. Give a phonemic transcription of the words below and then classify them into columns. Note that there is one word which does not fit in any column. liar joyous shower

layer sour hire

higher fire quiet

quite mower soya

towel gayer widower

19. Transcribe the following sentences: 1. She was pretty busy in her leisure time serving tea to English teachers. 2. She had the leather jacket cleaned but kept the sweater on. 3. My aunt asked the man who had put the apples in the basket and he answered back. 4. My uncle and my younger cousin ran past me. 5. The monk did not have any money for the bus to Gloucester. 6. My tongue and my stomach would love that honey. 20. Homophones. The following transcriptions represent a pair of homophones (unless otherwise stated). Give their spellings. 1. /nɒt/ 6. /ruːt/ 11. /stiːl/

2. /fɔːt/ 7. /luːz/ 12. /kɔːt/

3. /ɜːn/ 8. /wʊd/ 13. /ˈkɜːnl/

4. /juː/(3) 9. /kɔːs/ 14. /wɔːn/

5. /θruː/ 10. /wɒt/ 15. /kruːz/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:06 PM

Chapter 4

4 Consonants 4.1 Introduction This chapter describes RP consonants based on the acoustic, auditory and articulatory features (basically place and manner of articulation, position of the vocal folds, as well as force or energy of articulation) previously detailed in Chapters 1 and 2 (see under § 1.2, 2.3.2 and 2.4.3). As already noted, acoustically, consonants are characterised by irregular formant patterns, and articulatorily, they differ from vowels in that their production involves some sort of impediment or stricture in the VT, which is not present in vowel articulation. Here, in line with other reference books (Gimson 1980, 1984, 1994; Cruttenden 2008, 2014; Colina 2009), consonants are classified into two main groups, obstruents and sonorants, according to their sonority, that is, the correlation that exists between the degree of openness of the VT that is necessary for their articulation and the relative loudness with which they are perceived (see § 1.3.3.2 for details on the sonority hierarchy principle). Obstruents include three kinds of consonants that are produced with different types of obstruction to the airflow having resonance above that point of constriction (see fn 3). In the case of plosives the stricture impedes the airflow through nose or mouth; in fricatives the stricture causes friction; and in affricates there is a combination of both articulations. Articulatorily, obstruents show a distinctive opposition between fortis and lenis types; and acoustically they are typically associated with a noise component and with irregular formant structure patterns that show higher frequency. Phonologically, they tend to be “non-syllabic” or marginal in the syllable. Sonorants, on the other hand, are articulatorily defined as frictionless sounds that are produced with a relatively free, or unimpeded, oral or nasal airflow, having a vocal fold position such that spontaneous voicing is possible (see fn 4). They resonate throughout the VT. To this group belong four groups of consonants differing in manner of articulation, nasals, approximants (including glide consonants and liquids), taps and trills, which share many phonetic characteristics with vowels. Acoustically sonorants are characterised by the absence of a noise component and by homogeneous formant structure patterns that are lower in frequency. Some sonorants (nasals and liquids) may also be syllabic (see § 1.3.3.3).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

Introduction

151

As already explained, consonants are identified according to three-term VPM labels related to: (1) Voicing, i.e. whether or not the vocal cords vibrate in their articulation (voiced-voiceless) (see § 2.3.2.4). (2) Place of articulation, alluding to where the obstruction or stricture occurs with regard to eleven possible locations in the VT (bilabial, labio-dental, dental, alveolar, post-alveolar, palato-alveolar, palatal, velar, uvular, pharyngeal, and glottal) (see § 2.3.2.2); and (3) Manner of articulation, considering the type of obstruction that exists in the VT: broadly, plosives, affricates and nasals involve a complete closure of the oral tract; taps and trills are articulated with intermittent closures; laterals are produced with a partial oral closure; and fricatives need a stricture of close approximation (see § 2.3.2.1). Consistent with the contrastive standpoint justified in Chapter 3, section 4.2 compares the consonant systems of RP and PSp, focusing on the five groups that may be established according to their sonority and manner of articulation: plosives, fricatives, affricates − which make up the obstruent class −, nasals and approximants − which constitute the sonorant class. The members of each group are characterised along nine parameters: (1) IPA symbols, giving the names of the phonetic symbols related to each consonant phoneme. (2) Identification of RP consonants according to the VPM labels. (3) Allophones, listing the main allophonic realisations with their corresponding IPA symbols and meanings. (4) Description, describing the articulation of RP consonants. (5) Environment and main allophonic realisations, offering a description of the main contexts of appearance of RP consonants, as well as an explanation of their most relevant allophonic realisations, which are recapitulated in Chapter 5. (6) Spellings (with transcribed examples), listing the spellings of RP consonants in decreasing frequency and noting the most remarkable exceptions (if any). (7) Regional and social variants, mentioning the most important alternative pronunciations of consonants in RP and other related accents. (8) Comparison with Spanish and advice, where each RP consonant or consonant group is contrasted with the equivalent sounds in PSp providing pronunciation tips that are essential for intelligibility or otherwise may be relevant to the imitation of a native-like accent.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

152

Consonants

(9) Further practice, which includes recorded (and transcribed) examples (extracted from our Sound Bank). The reader is referred also to the Audio Units (exercises) to be listened to and practised. Remember that all the listening material is available in the companion EPSS Multimedia Lab.

4.2 A comparison of English and Spanish consonants The RP consonant system (annotated in bold type in Table 18 below) consists of twenty-four phonemes: six plosives /p t k b d ɡ/, nine fricatives /f v θ ð s z ʃ ʒ h/, two affricates /tʃ dʒ/, three nasals /m n ŋ/ and four approximants /j w l r/. The voiceless velar fricative [x] and the glottal plosive [Ɂ] are excluded from Table 18 because the former occurs exceptionally in some Scottish pronunciations of some words such as loch [lɒx] (vs. [lɒk]), whereas the glottal stop is not phonetically distinctive in RP, nor does it exist in the phonetic inventory of PSp. It will be explained that in RP [Ɂ] is mostly used as a reinforcement for vowels or as an allophone of plosives (see § 5.2.8). It should also be clarified that [w] appears twice in Table 18 to reflect its double labial-velar articulation, the former being placed between brackets due to its secondary nature (see § 4.2.5.2). As already remarked, only obstruents participate in the fortis /p t k f θ s ʃ tʃ/ – lenis /b d ɡ v ð z ʒ dʒ/ opposition,60 which is an important one for the repercussion it may have on the articulation of neighbouring sounds basically in terms of length and voicing (see § 2.3.2.1). The most remarkable are the two effects produced by fortis consonants. One is the reduction in length of previous vowels and sonorants (pre-fortis clipping, see § 2.3.1.4, 2.3.2.4, 5.2.1). As an illustration, compare rope [rəˑʊp], hurt [hɜˑt], leak [liˑk], sent [senˑt], self [seɫˑf], with reduced pre-fortis vowels and sonorants, with robe [rəːʊb], heard [hɜːd], league [liːɡ], send [senːd], selves [ˈseɫːvz], which have fully long vowels and sonorants. The other effect exerted by fortis plosives is the devoicing of a subsequent sonorant in consonantal clusters, as in e.g. quiet [ˈkw̥ aɪət] try [tɹ̥aɪ], play [pl ̥eɪ], pure [pj ̥ ʊə], where the four approximants are devoiced because they follow a fortis consonant in a stressed syllable (for more details see § 5.2.2). Now examining the phonemic oppositions that exist between the consonant systems of English and Spanish, it will be seen from Table 18 that while RP has twenty-four distinctive consonants, PSp only has nineteen (see also Table 4 in § 5.1), although these totals do not include allophonic or dialectal variations within either language (see § 1.3.1 and 5.2). Both languages have seven sonorant 60 /h/ constitutes a special case because it does not participate in the fortis-lenis opposition, like sonorants, but it lacks the voicing feature typical of the sonorant group.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

OBSTRUENTS Obstruction to the airflow; voiced or voiceless

s

s

z

d

ʃ

ʤ

ʒ

Midway between palatals and alveolars

Palatoalveolar

ʧ

ð

t

Postalveolar

PSp

θ

θ

d

Tongue tip against alveolar ridge

Alveolar

CORONAL61

ʧ

v

t

Tongue blade against upper teeth / between teeth

(Inter)Dental

RP

f

PSp

b

b

f

p

p

Lower lips and upper teeth

Two lips together

RP

PSp

RP

Labiodental

Bilabial

LABIAL

ʝ

Tongue blade towards hard palate

Palatal

k

k

x

ɡ

ɡ

Tongue body against velum

Velar

h

Between vocal cords

Glottal

61 ‘Coronal’ is a articulatorily cover-term for alveolar, dental and palato-alveolar consonants referring to sounds that are produced with the blade of the tongue raised from its neutral position. Its opposite, ‘non-coronal’, alludes to sounds that are articulated with the tongue blade in neutral position, as in labial and velar consonants (Crystal 2008: 117).

Affricates Begin as plosives; but end as fricatives

Fricatives Narrow opening friction

Plosives ðoral stopsÞ 3 stages: closing; compression release

Manner of articulation + Language

Place of articulation %

Table 18: The distinctive consonants of RP and PSp

A comparison of English and Spanish consonants

153

SONORANTS Greater degree of resonance; all voiced

Manner of articulation + Language

Place of articulation %

Table 18: (Continued)

Trill

Taps

Approximants No closure or friction

Nasals Velum lowered; airflow through nose

Glides

Liquids

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

PSp

RP

PSp

RP

l

n

n

Tongue tip against alveolar ridge

PSp

Tongue blade against upper teeth / between teeth

Alveolar

l

(w)

m

m

Lower lips and upper teeth

Two lips together

(Inter)Dental

RP

PSp

RP

PSp

RP

Labiodental

Bilabial

LABIAL

r

ɾ

r [ɹ]

Postalveolar Midway between palatals and alveolars

Palatoalveolar

CORONAL61

ʎ

j

ɲ

Tongue blade towards hard palate

Palatal

w

ŋ

Tongue body against velum

Velar

Between vocal cords

Glottal

154 Consonants

A comparison of English and Spanish consonants

155

consonants, RP /m n ŋ l r w j/ and PSp /m n ɲ l ʎ ɾ r/, whereas the rest are obstruents: seventeen in RP /p b t d ɡ k f v θ ð s z ʃ ʒ h ʧ ʤ/ and twelve in PSp /p b t d g k f θ s ʝ x ʧ/. Moreover, while both RP and PSp have six plosives and three nasals, they differ as to the number of phonemes consigned in the other manners of articulation, as summarised in Table 19. Table 19: Differences between the consonantal phonemic systems of English and Spanish English 24 consonants

Spanish 19 consonants

6 plosives /p t k b d ɡ/

6 plosives /p t k b d ɡ/

9 fricatives /f v θ ð s z ʃ ʒ h/

5 fricatives /f θ s ʝ x/

2 affricates /tʃ ʤ/

1 affricate /tʃ/

3 nasals /m n ŋ/

3 nasals /m n ɲ/

1 lateral /l/

2 laterals /l ʎ/

3 approximants /r w j/62

1 flap /ɾ/ 1 trill /r/63

In addition, the tables reveal that only eleven RP consonant phonemes have equivalent ones in PSp: /p b k ɡ f θ s tʃ m n l/. The other thirteen RP consonants have no exact equivalent phonemes in Spanish, /t d ʤ v ð z ʃ ʒ h ŋ r j w/. For this reason particular attention should be paid to these sounds as they can project possible phonemic difficulties for SSLE, as explained in Chapter 3, even though some of them may still appear in the phonetic or allophonic inventories of some dialects of Spanish. Nine belong to the obstruent category (see § 8.2.2). Firstly, although both languages have six plosives, /t d/ are dental in PSp, but alveolar in RP. Secondly, Spanish has only the voiceless affricate /tʃ/, while English has one pair, voiceless /tʃ/ (church /tʃɜːtʃ/) and voiced /ʤ/ (bridge [brɪʤ]). The other six obstruents that only occur in English belong to the fricative paradigm, /f v θ ð s z ʃ ʒ h/, which contains four voiceless-voiced pairs, as against only five single fricatives in PSp

62 Although here the wide-spread practice is adopted to annotate RP post-alveolar approximant as /r/, the IPA symbol that denotes this characterization is [ɹ] “Latin small letter turned R”, and so the latter will be used in narrow transcriptions to represent the post-alveolar approximant. 63 The IPA symbol /r/ stands for an alveolar trill, which corresponds to the description of the consonantal phoneme found in Spanish (as in perro ‘dog’), although other symbols have also been used to represent it such as /r̄/ or /R/ (mostly to denote all r types).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

156

Consonants

/f θ s ʝ x/: the voiced labio-dental /v/ (visit /ˈvɪzɪt/), dental /ð/ (rather /ˈrɑːðər/) and alveolar /z/ (zoo /zuː/) fricatives, both the voiced /ʒ/ (vision /ˈvɪʒn/) and unvoiced /ʃ/ (cash /kæʃ/) palato-alveolar fricatives, and the voiceless glottal fricative /h/ (ham /hæm/). The latter should not be confused with the spelling in Spanish, which is always silent, and so compare e.g. hola ‘hi’ /ˈola/ with hello /həˈləʊ/. The remaining four RP consonant phonemes that cannot be found in the phonemic inventory of Spanish belong to the sonorant group: the velar nasal /ŋ/ (thing /θɪŋ/), the postalveolar approximant /r/ [ɹ] (rise /raɪz/ [ɹaɪz]), and the glide consonants /j w/, which are here regarded as (semi-)consonants in RP, but as allophones of the vowels [i] and [u] in Spanish (see § 4.2.5.3 and 4.2.5.4). PSp, on the other hand, has six consonant phonemes that do not occur in RP: the palatal nasal /ɲ/ (caña /ˈkaɲa/ ‘cane’, close to /nj/ as in knew), the voiceless velar fricative /x/ ( jamón /xaˈmon/ ‘ham’), which comes close to RP /h/ but has a different place of articulation, the voiced palatal fricative /ʝ/ (yeso /ˈʝeso/ ‘plaster’) and the palatal lateral approximant /ʎ/ (llama /ˈʎama/ ‘flame’, close to /lj/ as in in million), the latter two resembling RP /ʒ/ and /ʤ/ but with a different place and manner of articulation, respectively. Lastly, while in Spanish there are two r types that have phonemic status, the alveolar tap /ɾ/ (pero ‘but’) and the alveolar trill /r/ (perro ‘dog’), in English [ɾ] is an allophone of /r/ (very [ˈveɾi]) and /t/ (butter [ˈbʌɾə]). Also relevant to the comparison between the consonant systems of English and Spanish are the following points (Fry 1947; Delatre 1965: 41; Navarro Tomás 1968; Monroy Casas 1980: 75; 2012; Finch and Ortiz Lira 1982: 62-65; Helman 2004; Swan and Smith 2001): (1) Generally speaking, Spanish consonants are shorter and are articulated with less muscular tension than their English counterparts, and are usually not so strongly released; as a result, variation of length of the consonant itself and of the preceding vowel is not used in Spanish as a cue to identify consonant quality as it occurs in English (bead [biːd] vs. beat [biˑt]). (2) Consonants in both English and Spanish show similar oscillations in place of articulation depending on their phonetic contexts of occurrence (see § 5.2.7). This is particularly so in the case of alveolars, but it also applies to velars and /m n/, which may have pre- or post-alveolar, pre- or post-velar, and labio-dental realisations ([ɱ]) if followed by /f v/, respectively. Likewise, in Castilian Spanish, though not in Latin American Spanish, /l n/ can have dental allophones [ ]̪ , if they precede /θ ð/, as in calzado ‘footwear’, as it occurs in English. (3) Turning to frequency counts, consonants occur more frequently in English than in Spanish, /s z/ and /k/ showing top frequency in the former, but

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

(4)

(5)

(6)

(7)

157

in the two languages the most common places of articulation are alveolar followed by velar. Note that this tendency is more skewed in English because seven consonant phonemes are articulated in the (post)alveolar region, as against only three in Spanish. Considering distribution, it should be noted that only six consonants have a restricted distribution in RP: /ʒ h ŋ r w j/. Both /ʒ/ and /ŋ/ are barred from word initial position, whereas /h r w j/ do not appear word finally. All the other consonants can occur word initially, medially and finally. In Spanish, on the other hand, consonants show a more restricted distribution as they are more frequently found in initial than in final positions, the latter being restricted to /d s n l/ and flapped r. As a result, speakers of Spanish may fail to pronounce final consonants accurately or strongly enough due to the higher number of consonants which may occur in that position in English. Especially problematic may be those English words ending in a consonant that does not exist or is barred from that position in Spanish. In terms of syllable structure, we have seen that while in English onsets and codas may have up to three and four consonants, respectively (spray /sprei/, sixths /sɪksθs/), Spanish exhibits a maximum of two consonants in both slots (trans-cribir ‘transcribe’) (see § 1.3.3.1). An examination of the spellings of consonants shows that English has over one hundred and twenty spellings for its inventory of twenty-four phonemes, whereas Spanish has twenty eight spellings to represent nineteen (see Chapter 7). Lastly, considering accent variation, in English it is vowels that vary the most from one regional variety to another, but in Spanish it is mainly consonants. In both languages, however, consonants have more allophonic variants than vowels.

Speakers of Spanish should make selective use of distinctive cues when learning English consonants taking into account their different L1 phonemic and phonetic inventories as they may give rise to pronunciation problems. Moreover, every effort should be made to accurately apprehend and reproduce the differences so far described, the most outstanding involving (1) phonemes that are common to English and Spanish but have different realisations (e.g. fortis plosives), (2) sounds which are phonemes in one language but allophones in the other (e.g. English /ð z/ vs. Spanish [ð z]), and (3) phonemes that occur only in English and therefore should be learnt anew by speakers of Spanish. The following are specific areas of difficulty and common replacements/mispronunciations detected in the pronunciation of SSLE, which are further discussed in the subsequent sections in connection with the particular phonemes ascribed to each

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

158

Consonants

consonant group (Sedláčková 2010; Estebas Vilaplana 2009; Helman 2004; Gallardo del Puerto 2005; Finch and Ortiz Lira 1982; Coe-Guerrero 1981): (1) /p/ tends to be pronounced as unaspirated and as a result it may sound almost like [b], which can result in miscommunication. The same is applies to /k/ (which may be confused with [g]) and /t/ (which may be replaced by [d]). (2) The /b/ – /v/ contrast that exists in English can be problematic. Whereas in English /b/ is always a stop, SSLE tend to pronounce [β], the approximant allophone of /b/, between voiced sounds, and also very frequently in replacement of RP /v/, especially in word-initial position. Moreover, in word-final position RP /b/ may be devoiced to [p] (as when cub is pronounced [kʌp] instead of /kʌb/). This replacement may similarly take place in the case of word-final /ɡ/ (which may become [k]) and /d/ (which may be realised as [t]). (3) In addition to being unaspirated, RP /t/ tends to be pronounced as dental (instead of alveolar) to match Spanish /t/. For the same reason, dental realisations of /d/ are also frequent, causing confusion with (and replacement by) [ð]. (4) Since the phoneme /ð/ does not exist in Spanish, it is often replaced by /d/. (5) RP /z/ is frequently pronounced as a voiceless fricative [s], or even as [θ] if pronunciation is misled by the common grapheme . (6) Due to its absence in Spanish, RP /ʃ/ is often replaced by [s] or [ʧ]. But, if the EFL learner knows how to pronounce the phoneme, s/he may overuse it and replace RP /ʧ/ by [ ʃ ] in some contexts. (7) RP /ʒ/ may be replaced by [ ʃ], [ʧ] or even [s]. (8) RP /h/ may be dropped, or otherwise it may be pronounced with a hissing sound or be replaced by Spanish [x]. (9) RP /ʤ/ tends to be confused with and replaced by [ ʝ], [ʎ] and [j], or even by [ʧ]. (10) In word-final positions RP /m/ may be replaced by [n] or [ŋ]. (11) As RP /ŋ/ does not occur in Spanish, it may be replaced by [n], [nɡ] or [nk]. (12) Learners may have trouble pronouncing dark l [ɫ]. (13) While in RP the post-alveolar approximant [ɹ] is a continuant sound, SSLE usually replace it by the Spanish flap [ɾ] or trill [r]. (14) RP /j/ may be confused with (or it can replace) [ʤ], as mentioned in (9) above, but due to the influence of Spanish it may also be mispronounced as a voiced palatal fricative [ ʝ] or a palatal lateral [ʎ], instead of as a palatal approximant. (15) RP /w/ may be mispronounced as a [b], or quite frequently, [ɡ] may be inserted before the sound itself.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

159

4.2.1 Plosives In English there are six plosive consonants, /p b t d k ɡ/, which can be also referred to as pulmonic, egressive, oral stops. Their articulation involves three successive stages (see also e.g. Gimson 1984: 152–171; Gimson 1994: 139–156; Roach 2005: 32–35; Collins and Mees 2003: 149–163; Collins and Mees 2009: 81– 87; Cruttenden 2014: 162–185): (1) Closing, closure or approach stage, during which the active articulator raises towards the passive articulator in order to form a complete closure or obstruction to the airstream. This is the defining characteristic of stops; if there is no closure, then there is no stop. (2) Hold or compression stage, during which, the lung air cannot go out of the mouth as it is stopped at the point where the two articulators meet. When we pronounce plosives the soft palate is in its raised position, and as a result the lung air cannot be released through the nose either (unless a nasal consonant follows the plosive). The vocal folds may or may not vibrate. (3) Release or explosion stage, during which the two articulators are abruptly separated so that the lung air goes out of the mouth producing a kind of explosion (hence the term plosive). If the vocal folds vibrate in the compression stage, they continue to do so in the release stage; but if stage (2) is voiceless, then stage (3) may either be also voiceless (aspirated) or it may be the onset of vocal fold vibration. There exists an off-glide, or transition, from the plosive to the following sound. In addition, the plosive paradigm can be further classified according to the three-term VPM series used to classify consonants: (1) Position of the vocal folds, either they do not vibrate in voiceless or unvoiced plosives /p t k/, or they do vibrate, when pronouncing the voiced ones /b d ɡ/. There exists allophonic variation involving different degrees of devoicing [b̥ d̥ ɡ̊] or voicing [p̌ t ̬ k̬] in the case of voiced and voiceless plosives, respectively. Devoiced realisations of voiced plosives occur in initial and especially in final positions (Did [d̥ɪd̥]); while fully voiced realisations, left unmarked in allophonic transcriptions, arise in word-medial, intervocalic position (labour [ˈleibə]. In the latter phonetic context voiceless plosives have voiced realisational variants (matter [ˈmæt̬ə] (see § 1.3.2 and 2.3.2.1). (2) Place of articulation: /p b/ are bilabial, the two articulators that are used when we pronounce them are the two lips, the lower lip, the active articulator, and the upper lip, the passive articulator. /t d/ are alveolar because the two articulators are the tip of the tongue, the active articulator, and the

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

160

Consonants

alveolar ridge, the passive articulator. Finally, /k ɡ/ are velar, the back of the tongue and the soft palate acting as the active and passive articulator, respectively. (3) Force or energy of articulation: /p t k/ are fortis or strong consonants because they are pronounced with more muscular force or energy than /b d ɡ/, which are lenis or weak consonants. As already remarked, long vowels and diphthongs are shortened or clipped when followed by a fortis consonant either within the word or in the following word. For example, in seek the vowel is shortened [siˑk], while in weed [wiːd], the vowel is fully long. This shortening helps to differentiate many word-pairs such as mate [meˑit] and made [meːid]. Three additional features must be taken into account to characterise the plosive group: release type, aspiration and their Voice Onset Time (VOT) values. Considering the type of release, plosives may have three different allophonic variants depending on their position in the word and on the consonant that follows them: (1) Non audible release, annotated in narrow transcriptions with a superscript high-tone mark [p˺ t˺ k˺ b˺ d˺ ɡ˺], is attributed to the allophones of so-called incomplete plosives in the sense that the third (release or explosion) stage is either non-audible or it does not exist. More common in the velar pair than in the bilabial or alveolar ones (Byrd 1992), incomplete plosives are restricted to two main contexts: word-finally, particularly before a silence and in rapid speech (road [rəʊd˺]), and in consonantal clusters, where incomplete plosives occur if they are followed by another plosive or by an affricate either within a word, both in the same or a different syllable, or between words, as shown in the following examples: act [æk˺t], picture [ˈpɪk˺tʃə], track tape [ˈtɹæk˺ ˈteip], big ball [ˈbɪɡ˺ ˈbɔːl], webcast [ˈweb˺kʰɑːst]. Accordingly, the articulatory steps that are necessary to pronounce, for instance, incomplete [b] in the [b˺k] sequence of the last example would be as follows. In the first stage, the two articulators, the two lips, come into contact in order to produce [b], the soft palate is in its raised position. In the second stage, the lung air cannot go out through the nasal passage and it is compressed in the mouth just behind the closure made by the two articulators, the back of the tongue is placed close to the soft palate. In the third stage, the articulators are separated and the lung air goes out of the mouth producing a kind of explosion. Two plosives are pronounced but there is only one explosion.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

161

In addition, cases of gemination (i.e. sequences of identical stops, as in top post, good day)64 and series of homorganic plosives (i.e. with the same place of articulation) that differ in voicing (top brand, that day, big coach) involve only one closing stage and one release stage with an approximately double-length compression stage. In the case of homorganic plosive clusters, the first plosive is more accurately termed unreleased and is annotated with a superscript up arrowhead [ˆ]: [ˈtʰɒpˆ ˈpʰəʊst=], [ˈgʊdˆ ˈdeɪ], [ˈtɒpˆ ˈbɹænd] [ðətˆ ˈdeɪ], [bɪɡˆ kəʊtʃ ]). Cues for the recognition of voiceless or voiced sounds are provided by onset or cessation of voice, if voiced, and, if voiceless, by the presence of aspiration and the shortening of previous sonorants. Finally, in the event of clusters with three plosives, the first two plosives are incomplete (or unreleased in homorganic plosive clusters). For in addition to the omission of the third stage of the first plosive, the central plosive has non-audible first and third stages (rubbed pork [rʌb˺d˺ pɔːk]), and if that slot is occupied by [p t k], then the second stage is manifested by a silence of certain duration (stopped begging [ˈstɒp˺t˺ ˈbeɡɪŋ]). As will be apparent from the transcribed examples, incomplete fortis plosives are always unaspirated because they lack the release stage when aspiration is produced. (2) Nasal release, annotated with a superscript lower-case N [pⁿ tⁿ kⁿ bⁿ dⁿ ɡⁿ], takes place when the plosive is followed by a nasal consonant, both within and between words, as illustrated in the following cases: topmost [ˈtɒpmməʊst], submerge [səbmˈmɜːdʒ], chutney [ˈtʃʌtⁿni], chicken [ˈtʃɪkⁿŋ̩], cheap mattress [tʃiːpm ˈmætɹɪs]. If the nasal is homorganic, what we do is just to keep the two articulators together in the oral chamber while lowering the soft palate so that the compressed air is liberated through the nasal passage. But if the nasal is not homorganic (big mall [bɪɡm mɔːl]), then the plosive is not released until the articulatory movements for the nasal consonants are performed (change of place of articulation in the oral chamber and lowering of the soft palate); otherwise, some oral release may be perceived. (3) Lateral release, transcribed with a superscript lower-case L [pˡ tˡ kˡ bˡ dˡ gˡ], occurs when a plosive is followed by a lateral consonant. The plosive is then said to be released laterally, both between and within words, either when 64 The only case where geminated letters are pronounced as double sounds is when they occur across words (can never /kən ˈnevə/) or when they correspond to two different grammatical categories within words: (1) prefix + word (midday /ˌmɪdˈdeɪ/), (2) word + suffix (leanness /ˈliːnnəs/), (3) compounds (hop-pole /ˈhɒp pəʊl/). Even though two sounds appear in transcription, most of the times double sounds are pronounced as single sounds, except that they last longer. An exception to this are the series consisting of two affricates, in which case the sounds are produced twice (rich church /ˈrɪtʃ ˈtʃɜːtʃ/ orange juice /ˈɒrɪndʒ dʒuːs/).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

162

Consonants

the plosive is followed by syllabic [ɫ ]̩ , as in little [ˈlɪtˡɫ ̩], or when [l] is initial in the next syllable or word, as in bad leg [bædˡ leg]. This means that the lung air goes out laterally because one or the two sides of the tongue are lowered to allow the air to escape to pronounce [l], the plosive contact remaining. A distinction can be made between “true lateral release”, occurring in homorganic plosive-nasal sequences [tˡl], [dˡl], where there is no intervening removal of the tongue contact on the alveolar ridge so that the air cannot escape centrally over the tongue, and the other laterally released plosives in [p b k ɡ] + [l] sequences, where the escape of air is lateral in the sense that the partial alveolar contact for [l] is made before or at the time of the release of the plosive (apple [ˈæpˡɫ ̩] fake leg [ˈfeɪkˡ ˈleɡ], glow [ɡˡləʊ]) (Cruttenden 2014: 172). Let us now turn to the other two features that characterise plosives: aspiration and VOT. Aspiration (annotated with superscript lower-case H) is an articulatory feature that may be present in the release stage of RP fortis plosives in a specific context. In Section 2.3.2.1 it was explained that, when occurring in stressed syllable-initial position, fortis plosives are aspirated [pʰ tʰ kʰ], i.e. they are articulated with a brief puff of air (Crystal 2008: 38–39), as in pan [pʰæn], tan [tʰæn] and can [kʰæn]. In all other contexts fortis plosives are either unaspirated [p= t= k=], after /s/ as in ski [sk=iː], or otherwise have a relatively weak aspiration, in unaccented syllables (police [p=əˈliːs]) and in both word and syllable final positions (tap [tʰæp=]). Notice, however, that when fortis plosives are followed by approximants, especially in syllable-initial stressed position, aspiration is manifested regardless of its degree as approximant devoicing [l ̥ ɹ̥ j ̥ w̥ ] (e.g. play [pl ̥eɪ], splendid [ˈspl ̥endɪd], try [tɹ̥aɪ], strange [stɹ̥eɪndʒ], obscure [əbˈskjʊə], quite [kw̥ aɪt=]) (see § 5.5.2). ̥ The acoustic correlate of aspiration, VOT, measures in milliseconds the interval between the release burst stage and the onset of voicing (Lisker and Abramson 1964; Ladefoged, 2001: 120; Cruttenden 2014: 164–165). Broadly, aspirated voiceless plosives [pʰ tʰ kʰ], as [pʰ] in pea [pʰiː], exhibit positive VOTs, or voicing lags (i.e. the vocal-fold activity starts after the release of the plosive); whereas unaspirated voiceless plosives [p= t= k=], as in ski [sk=iː], have a zero VOT (i.e. the plosive release and the onset of vocal-fold vibration are simultaneous). The VOT values for voiceless plosives have been found to increase the further back the place of articulation is from labial to velar (Docherty 1992; Volatis and Miller 1992). In contrast, fully voiced [b d g] and devoiced voiced plosives [b̥ d̥ g]̊ have long and short negative VOTs or voicing leads (i.e. the voicing starts before the release or explosion stage) respectively. These VOT differences of plosives are graphically represented in Figures 48 and 49, where

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

Figure 48: VOT values for RP plosives

Figure 49: Examples of illustrating the VOT values for RP plosives

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

163

164

Consonants

one can see the impact that VOT values have on the articulation of the subsequent sound. Differences reside mainly in the point at which voicing starts in relation to the release stage of plosives. (De)voiced and unaspirated voiceless plosives trigger non-delayed onsets to voicing: voiced plosives have “unbroken voicing” for which no VOT values can be specified (obey [əˈbeɪ]); devoiced lenis plosives have negative VOTs (bay [b̥eɪ]); and unaspirated voiceless plosives have zero VOTs (stay [st=eɪ]). Aspirated voiceless plosives, in contrast, with positive VOTs are followed by sounds showing delayed onsets to voicing (Tay [tʰeɪ]). Further details on VOT differences may be found in Sections 2.4.1 and 2.4.3, where the acoustic features of vowels and plosives are discussed. It was explained that fortis plosives have a voiceless silent interval, which is long if unaspirated, and is followed by a burst if aspirated. Voiced stops, in contrast, have generally a shorter closure during which there is a voice bar and the release burst is weaker and has no aspiration. More details about the allophonic variants of English plosives (e.g. place of articulation, lenition and fortition), as well as the effects they exert on the neighbouring sounds are offered in Chapter 5 (see § 5.2 and 5.3). To conclude this introduction to the plosive group, let us focus on the major differences and pronunciation difficulties that they may pose for speakers of Spanish (Finch and Ortiz Lira 1982: 62–65; Estebas Vilaplana 2009: 49–64; Cruttenden 2014: 173, 328). Learners can become aware of the different nature of RP phonemes by means of ear training and oral practice. As already advanced, particular attention should be paid to the aspiration of RP fortis plosives in syllable-initial stressed position because failure to do so can lead to intelligibility or otherwise can result in a foreign accent. While aspiration is the main distinguishing cue to differentiate [p t k] from [b d g] in initial acquisition stages by native speakers of English (Macken and Barton 1980), this feature may be considered as non-existent in Spanish, where the contrast between plosive pairs relies exclusively on the presence-absence of voice. Consequently, SSLE have a tendency to pronounce words like pen as *[pen] (without aspiration), instead of [pʰen] (with aspiration), which may be misunderstood by English listeners as Ben [ben] since they interpret lack of aspiration as a mark of voiced [b] (Stockwell and Bowen 1965; Monroy Casas 1980; Dale and Poms 1986: 81; Swan and Smith 2001; Catford 1987; Avery and Ehrlich 1992; Gallardo del Puerto 2005: 196–197; Sedláčková 2010: 59; Mott 2011; Gleason 2012). Likewise, if learners do not devoice sonorants after fortis obstruents (plosives and fricatives), an English ear may interpret, for instance, that a lenis rather than a fortis plosive is produced (e.g. glue, bride and dune instead of clue, pride and tune), leading to miscommunication or misunderstandings. In fact, VOT duration is one of the factors that have most influenced on the native English speakers’ perception of a foreign accent in SSLE’s pronunciation (Flege 1991; Flege and Eefting 1988). To

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

165

practice this feature, visual reinforcement can be obtained by holding a piece of paper in front of your mouth, which should move at the production of an aspirated fortis plosive. Aspiration of lenis plosives should be avoided. This feature has been traced in some speakers of Spanish resulting from hypercorrection and an effort to sound less Spanish. Other characteristics of plosive realisation that tend to be mastered last but should also be attended to by learners who aim at a near approximation to RP are: (1) the nasal and lateral release of alveolar plosives followed by homorganic [n] and [l], without any intervening [ʰ] or [ə]; (2) the production of incomplete plosives following other plosives or affricates; and (3) the devoicing of sonorants following a plosive. Although in Spanish plosives may also have a nasal release (e.g. [tn] [kn] in étnico ‘ethnic’ and técnico ‘technical’ ‘technician’) or a lateral release (e.g. [tˡl] [bˡl] in atlas ‘atlas’ and sable ‘sable’), the production of alveolar plosives with nasal or lateral release within the same syllable may be potentially complicated. Learners are advised to keep their tongue against the alveolar ridge so as to avoid the insertion of a vowel. The production of incomplete plosives may also be difficult for speakers of Spanish because in this language plosives do not normally occur together (but notice such series as [p˺t] and [k˺t] in apto and acto), and they never appear in final position. In addition, SSLE should be careful not to fricativise English plosives in contexts in which this is not allowed in English. Whereas in English [b d g] are largely realised as plosives (albeit the affricated allophones discussed in Section 5.2.8.1), in Spanish they only have a plosive realisation before a pause and after nasals (e.g. [mb] cambio ‘change’) and /l/ (e.g. Maldivas), being fricativised elsewhere. Thus, if a lenis plosive intervenes in a plosive cluster or is followed by a sonorant, then it is realised as a fricative in accordance with the commonest tendency of Spanish to realise voiced plosives [b d g] as corresponding voiced fricatives or approximants [β ð ɣ] in most contexts (e.g. [ðɣ] Edgardo, [βð] abdicar ‘abdicate’, [ðm] admirar ‘admire’, [ɣm] dogma ‘dogma’) (Stockwell and Bowen 1965; Navarro Tomás 1966; Quilis and Fernández 1982; Finch and Ortiz Lira 1982; Mott 2011). As these fricative sounds have an allophonic, but not phonemic status in Spanish, it may be the case that in English sequences where there is an initial fricative and an intervocalic plosive (they do /ðeɪ ˈduː/) SSLE reverse the order of these two sounds (*[deɪ ˈðu]) which should be avoided in order not to sound foreign (Gallardo del Puerto 2005: 197). 4.2.1.1 Bilabial plosives /p b/ IPA Symbols p Lower-case P. b Lower-case B.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

166

Consonants

Identification /p/ Fortis voiceless bilabial plosive. /b/ Lenis voiced bilabial plosive. Allophones /p/ [pʰ] ‘aspirated’ [p=] ‘unaspirated’ [ʔ] ‘glottalised’ [p̌] ‘voiced’ [p˺] ‘non-audible release’ [pˆ] ‘unreleased’ [pⁿ] ‘nasal release’ [pˡ] ‘lateral release’ [p̪] ‘labio-dental’ [p̠] ‘more retracted’ [pʷ] ‘labialised’ [pʲ] ‘palatalised’ [ɸ] ‘voiceless bilabial fricative’

/b/

[b̥] ‘devoiced’ [b˺] ‘non-audible release’ [bˆ] ‘unreleased’ [bⁿ] ‘nasal release’ [bˡ] ‘lateral release’ [b̪] ‘labio-dental’ [b̠] more retracted’ [bʷ] ‘labialised’ [pʲ] ‘palatalised’ [β] ‘voiced bilabial fricative’

Description When we pronounce /p/ the two lips form a closure and the lung air is stopped behind it; the soft palate is in its raised position, so that there are two closures. In the release stage, the two articulators are separated and the lung air goes out of the oral cavity producing a kind of explosion. The vocal folds do not vibrate, they are apart. /p/ is produced with more muscular energy than its voiced counterpart, and it shortens the vocalic sounds that precede it (beep [biˑp]). /b/ is pronounced in the same way as /p/ but with less muscular energy, and the vocal folds may vibrate for all or part of the compression stage. Aspiration does not take place in the production of this sound, nor does it affect the length of the preceding vocalic or sonorant sound (urb [ˈɜːb]). See the animations and videos under RP /p/ and /b/ (compared with Spanish /p b/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP /p b/ can appear in all possible contexts. Initially in a stressed syllable /p/ is aspirated [pʰ] (pan [pʰæn], appear [əˈpʰɪə]). Finally in the word, it loses some of that aspiration [p=] (tap [tʰæp=]), and when preceded by /s/ it is unaspirated [p=] (spider [ˈsp=aɪdər]). Considering voice, between voiced consonants /p/ may be voiced [p̌ ] (April [ˈeɪp̌ɹəl], upon [əˈp̌ɒn]) and /b/ is fully voiced (rabbit [ˈɹæbɪt],

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

167

rubber [ˈɹʌbər], husband [ˈhʌzbənd]), but the latter loses some of its voicing when it occurs in initial position (big [b̥ɪg]) and is completely devoiced in final position (rob [ɹɒb̥]). /t/ may also be glottalised [ʔ] word-medially (supper [ˈsʌʔə]) and in homorganic consonant clusters that occur word-finally or at close-knit word-boundaries (soap powder [ˈsəʊʔ ˈpaʊdə] (see Regional variants and Section 5.2.8.2). In consonantal clusters, both plosives are incomplete, with non-audible release [p˺ b˺], if followed by another plosive or by an affricate (stopped [stɒp˺t], rubbed [rʌb˺d]), or unreleased [pˆ bˆ], if followed by a homorganic plosive (top brand [ˈtɒpˆ ˈbɹænd]). They may also be produced with nasal release [pⁿ bⁿ], nasal plosion being heard, when followed by a nasal consonant (shipmate [ˈʃɪpmmeɪt], keep narrating [kiːpⁿ nəˈreɪtɪŋ], submit [səbmˈmɪt]). The plosion will be lateral and they will have a lateral release [pˡ bˡ], if followed by a lateral consonant ( play [pˡleɪ], blue [bˡluː]). Besides, in the above examples and elsewhere sonorants are devoiced after /p/ (play [pl ̥eɪ]). Both stops are often produced with a labio-dental [p̪ b̪] rather than a bilabial closure when they are followed by a labio-dental fricative /f v/ (cupful [ˈkʌp̪fʊl], obviate [ˈɒb̪vɪeɪt]). Likewise, more retracted realisations occur, especially in consonantal clusters, when a following sound has a more retracted place of articulation, as in price [p̠ɹaɪs] and brush [b̠ɹʌʃ ] (see § 5.2.7). In addition to the glottalised and affricated variants discussed in Section 5.2.8, other allophonic realisations show different degrees of lip-rounding (bog [bʷɒɡ) or lip-spreading ( pea [pʲiː]) in the vicinity of rounded and spread vowels respectively (see § 5.2.3). Spellings Table 20: The spellings of RP /p b/ /p/

/b/

hop /hɒp/, hope /həʊp/ hoping /ˈhəʊpɪŋ/

rub /rʌb/, back /bæk/

hopping /ˈhɒpɪŋ/, supper /ˈsʌpər/

rubber /ˈrʌbər/, ebb /eb/

shepherd /ˈʃepəd/ Rare cases

hiccough /ˈhɪkʌp/

Rare cases silent

+ cupboard /ˈkʌbəd/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

168

Consonants

The letter p is silent when it occurs before n, s or t in several words of Greek origin: pneumonia /njuːˈməʊnɪə/ pseudo /ˈsjuːdəʊ/

psychology /saɪˈkɒlədʒi/) pterodactyl /ˌterəˈdæktɪl/

In addition, p is silent in cupboard, where it is followed by pronounced /b/ as already noted in Table 20, and in the following words: corps /kɔː/ coup /kuː/

raspberry /ˈrɑːzbəri/ receipt /rɪˈsi:t/

In spellings p may or may not be silent, and so alternative pronunciations occur with and without it (e.g. attempt /əˈtempt/ – /əˈtemt/, tempt /tempt/ – /temt/, empty /ˈempti/ – /ˈemti/). The letter b, on the other hand, is silent in and combinations, mostly in final position, as in: climb /klaɪm/ comb /kəʊm/ crumb /krʌm/ lamb /læm/ numb /nʌm/ plumb /plʌm/ thumb /θʌm/ tomb /tuːm/

plumber /plʌmər/ Woolacombe /ˈwʊləkəm/

debt /det/ doubt /daʊt/ subtle /ˈsʌtl ̩//

Regional and social variants Both the amount of aspiration given to /p/ and the degree of devoicing associated with /b/ varies between speakers (Cruttenden 2014: 175). In some dialects (Southern Irish, Highland Scottish) /p/ shows stronger aspiration, while in others it has only little aspiration (Lancashire). In RP it is increasingly typical to reinforce word-final fortis plosives with a glottal closure (see § 5.2.8.2 glottal reinforcement). Some speakers may also devoice /b/ in intervocalic positions. Comparison with Spanish and advice There are two issues that Spanish speakers should attend to in this group: the distinction between /p/ and /b/ in initial and final positions, and the fully voiced realisation of intervocalic /b/ (Cruttenden 2014: 175; Estebas Vilaplana 2009: 49–62). Whereas in Spanish /b/ is always voiced, in English it is devoiced [b̥] in initial and final positions, and so in these two contexts [b] sounds close to [p]. Accordingly, in order to not sound foreign and not to affect the intelligibility

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

169

of the two bilabial plosives, it is essential that SSLE pronounce [pʰ] with aspiration in stressed syllable-initial position (appear [əˈpʰɪə]); otherwise, as already explained, they will sound foreign and communication problems may arise because the absence of aspiration suggests /b/ to an English ear, and so words such as pen may be misunderstood as Ben, pin as bin and the like. Furthermore, the main cue to distinguish /p/ and /b/ when they precede /l r w j/, as in plead – bleed, pray – bray, is the devoicing of the latter only when they follow accented syllable initial aspirated /p/, giving the following contrasts [pl ̥iːd] – [bliːd], [pɹ̥eɪ] – [bɹeɪ] (Dale and Poms 1986; Swan and Smith 2001; Avery and Ehrlich 1992; Gleason 2012; Cruttenden 2014: 328–329). In final positions, no problems seem to arise because speakers of Spanish tend to pronounce [b] as [p]. In this context, the major distinguishing cue that native English speakers use is not aspiration because /p/ tends to be unaspirated, but rather the length of the preceding vowel: it is shorter if followed by [p=] (rope [ɹəˑʊp=]) than when the following sound is devoiced (robe [ɹəːʊb̥]). Nevertheless, SSLE should take special care not to pronounce word-final /b/ as [f] in such words as pub *[paf], which may occur as an interference from the tendency of Spanish to fricativise final plosives (Gallardo del Puerto 2005: 200). Another aspect that causes a strong foreign accent is the tendency that Spanish speakers have to produce a voiced bilabial fricative [β] in intervocalic position, as in robo ‘theft’ [ˈroβo], when this is the only environment in which RP /b/ is a fully voiced bilabial stop (robber [ˈɹɒbər]). To overcome this pronunciation problem, Spanish learners are advised to produce English intervocalic [b] as the initial sound of the word bola ‘ball’. Further practice AI 4.1 The spellings of RP /p/. Further practice pack /pæk/ simple /ˈsɪmpl ̩/ cap /kæp/

play /pleɪ/ pepper /ˈpepə/ tap /tæp/

taper /ˈteɪpə/ hopping /ˈhɒpɪŋ/

Perhaps you can pay for the shopping? /pəˈhæps ju kən peɪ fə ðə ˈʃɒpɪŋ/ She plays the piano perfectly. /ʃi pleɪz ðə pɪˈænəʊ ˈpɜːfɪktli/ Peter Piper picked a peck of pickled peppers. /ˈpiːtə ˈpaɪpə pɪkt ə pek əv ˈpɪkl ̩d ˈpepəz/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

170

Consonants

AI 4.2 The spellings of RP /b/. Further practice back /bæk/ symbol /ˈsɪmbl ̩/ cab /kæb/

bin /bɪn/ abroad /əˈbrɔːd/ rubber /ˈrʌbə/

club /klʌb/ rabbit /ˈræbɪt/

Betty brought bags of beautiful pebbles. /ˈbeti ˈbrɔːt bæɡz əv ˈbjuːtəfl̩ ˈpebl ̩z/ My brown rabbit eats cabbage and bananas. /maɪ braʊn ˈræbɪt iːts ˈkæbɪdʒ ənd bəˈnɑːnəz/ We’d better book a cab. /wid ˈbetə bʊk ə kæb/ Betty and Bob brought back blue balloons from the big bazaar. /ˈbeti ənd bɒb ˈbrɔːt bæk bluː bəˈluːnz frəm ðə bɪɡ bəˈzɑː/ 4.2.1.2 Alveolar plosives /t d/ IPA Symbols t Lower-case T. d Lower-case D. Identification /t/ Fortis voiceless alveolar plosive. /d/ Lenis voiced alveolar plosive. Allophones /t/ [tʰ] ‘aspirated’ [t=] ‘unaspirated’ [ʔ] ‘glottalised’ [ɾ] ‘alveolar tap’ [t ]̬ ‘voiced’ [t˺] ‘non-audible release’ [tˆ] ‘unreleased’ [tⁿ] ‘nasal release’ [tˡ] ‘lateral release’ [t̪] ‘labio-dental’ [t̠] ‘retracted’ ‘post-alveolar’ [tʷ] ‘labialised’ [tʲ] ‘palatalised’ [s] ‘voiceless alveolar fricative’

/d/

[d̥] ‘devoiced’ [d˺] ‘non-audible release’ [dˆ] ‘unreleased’ [dⁿ] ‘nasal release’ [dˡ] ‘lateral release’ [d̪] ‘labio-dental’ [d̠ ] ‘retracted’ ‘post-alveolar’ [dʷ] ‘labialised’ [dʲ] ‘palatalised’ [z] ‘voiced alveolar fricative’

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

171

Description When we pronounce /t/, the soft palate is in its raised position and the tip of the tongue comes into contact with the alveolar ridge so that a closure is formed and the lung air is blocked at that point. The air escapes upon the sudden separation of the alveolar closure producing a kind of explosion. The vocal folds do not vibrate, but are wide apart. /t/ is produced with more muscular energy than its voiced counterpart and it reduces the length of preceding vocalic and sonorant sounds (beat [biˑt]). /d/ is articulated in the same way as /t/, but with less muscular energy, and the vocal folds may vibrate for all or part of the compression stage. Aspiration does not take place in the production of this sound, nor does it shorten the length of preceding vocalic or sonorant sounds (bead [biːd]). See the animations and videos under RP /t/ and /d/ (compared with Spanish /t d/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP /t d/ can appear in all possible contexts. In syllable-initial stressed position /t/ is aspirated [tʰ] (tea [tʰiː], attire [əˈtʰaɪə]). Finally in the word, it loses some of that aspiration [t=] (cat [kʰæt=]), and when preceded by /s/ it is unaspirated [t=] (stay [ˈst=eɪ]). Considering voice, between voiced sounds /t/ acquires some voicing [t ]̬ (attain [əˈt̬eɪn], spitting [ˈspɪt̬ɪŋ]) and /d/ is fully voiced (addition [əˈdɪʃn̩], London [ˈlʌndən], ladder [ˈlædər]), but the latter loses some of its voicing [d̥] when it occurs in initial position (day [d̥eɪ]) and is completely devoiced in final position (cod [kɒd̥]). In syllable-final positions, /t/ is commonly reinforced or replaced by a glottal closure (late [leɪʔt] [leɪʔ], bent [benʔt] [benʔ]) (for further details see § 5.2.8.2 on glottalisation and glottaling). In addition, in weakly accented contexts /t d/ are especially liable to affrication [ts dz] (time [ˈtsaɪm], dime [dsaɪm]) or even replacement by corresponding fricatives (important [ɪmˈpɔːtsn̩t] [ɪmˈpɔːsn̩t]) (see § 5.2.8.1 on affrication and fricativisation). In consonantal clusters, both plosives are incomplete, with non-audible release [t˺ d˺], if followed by another plosive or by an affricate ( football [ˈfʊt˺bɔːl], white chalk [ˈwaɪt ˈtʃɔːk], bad play [ˈbæd˺ ˈpleɪ], bed changes [ˈbed˺ ˈtʃeɪndʒɪz]), or unreleased [tˆ dˆ], if followed by a homorganic plosive (white doll [waɪtˆ dɒl], bad tan [ˈbædˆ tæn]). They may also be produced with nasal release [tⁿ dⁿ] and nasal plosion will be heard, when followed by a (homorganic) nasal consonant (curtness [ˈkɜːtⁿnəs], nutmeg [ˈnʌtᵐmeɡ], not now [nɒtⁿ naʊ], admit [ədᵐˈmɪt], red notebook [ˈredⁿ ˈnəʊtbʊk]). Lateral plosion will be heard if [l] follows, in which case alveolar plosives have a homorganic lateral release [tˡ dˡ] (at least [ətˡ liːst] middle [ˈmɪdˡɫ ̩), that is, only part of the alveolar obstruction is removed as the tongue-tip contact remains. Note also that in the above examples and elsewhere sonorants are devoiced after /t/ (try [tɹ̥aɪ]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

172

Consonants

The alveolar contact is the most sensitive to the place of articulation of a subsequent sound (see § 5.2.7), which is demonstrated by the ease with which alveolar consonants, particularly /t d/, may be elided in consonantal clusters (see § 5.4). Thus, alveolar stops may have post-alveolar realisations [ ]̠ , if followed by /r/, which is realised as a voiced fricative [ɹ̝] after /d/, but as a voiceless fricative after /t/, so that the two allophonic series are pronounced [d̠ ɹ]̝ [t̠ɹ]̥ (dry [d̠ɹaɪ], tray [t̠ɹe̥ ɪ]). Dental allophones [t̪ d̪] are produced before /θ ð/ (not those [ˈnɒt ̪ ðəʊz], hide this [ˈhaɪd̪ðɪs]). The lip position for the articulation of /t d/ will also be conditioned by that of the adjacent sounds. Thus, allophonic realisations show different degrees of anticipatory lip-rounding (too [tʷuː], twice [tʷwaɪs], door [dʷɔː], dwindle [ˈdʷwɪndl ̩]) or lip-spreading (tea [tʲiː], duty [ˈdʲjuːti]) in the vicinity of rounded and spread vowels or semi-vowels, respectively (see § 5.2.3). Spellings Table 21: The spellings of RP /t d/ /t/

/d/

vital /ˈvaɪtl ̩/, tell /tel/, cat /kæt/

mad /mæd/, date /deɪt/, medal /ˈmedl ̩/

hitting /ˈhɪtɪŋ/, attack /əˈtæk/

added /ˈædɪd/, Eddy /ˈedi/

65 Thames /temz/, Thomas /ˈtɒməs/ thyme /taɪm/, Thailand /ˈtaɪlænd/

washed /wɒʃt/, laughed /lɑːft/

opened /ˈəʊpənd/, bombed /bɒmd/

typed /taɪpt/

cigarette /ˌsɪɡəˈret/ silent + brought /ˈbrɔːt/ Rare cases

two /tuː/ silent + yacht /jɒt/

Rare cases silent + would /wʊd/

65 Note that the NP of spellings is /θ/ or /ð/ (see § 4.2.2.2), although they may also be silent (asthma /ˈæsmə/).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

173

Table 21 shows that the inflectional ending for past tenses and past participles of regular verbs ending in a voiced sound other than /d/ is pronounced /d/ (seized /siːzd/, played /pleɪd/); but if they end in a voiceless consonant other than /t/, then it is pronounced /t/ (packed /pækt/, stretched /stretʃt/). When the last sound of the stem of such verbal forms is a /d/ or a /t/, the inflectional ending is pronounced /ɪd/ (landed /ˈlændɪd/, waited /ˈweɪtɪd/). Adjectives ending in , on the other hand, follow the same pronunciation principles as those at work in the pronunciation of the verbs from which they derive: /ɪd/ or /əd/ after /t/ or /d/ (excited /ɪkˈsaɪtɪd/ – /ɪkˈsaɪtəd/), /d/ after voiced consonants other than /d/ (surprised /səˈpraɪzd/), and /t/ after voiceless consonants other than /t/ (embarrassed /ɪmˈbærəst/). There are a number of adjectives, however, that contravene these principles and have different pronunciations for the verbal and adjectival uses of the same forms, as shown below: Spelling aged blessed crabbed crooked dogged jagged learned ragged rugged -legged naked sacred wicked wretched

Adjective /ˈeɪʤɪd/ – /ˈeɪʤd/ /ˈblesɪd/ – /ˈblesəd/ /ˈkræbɪd/ – /ˈkræbd/ /ˈkrʊkɪd/ /ˈdɒgɪd/ /ˈʤægɪd/ /ˈlɜːnɪd/ /ˈrægɪd/ /ˈrʌgɪd/ /ˈlegɪd/ – ˈlegd/ /ˈneɪkɪd/ /ˈseɪkrɪd/ /ˈwɪkɪd/ /ˈretʃɪd/

Verb /eɪʤd/ /blest/ /kræbd/ /krʊkt/ /dɒgd/ /ʤægd/ /lɜːnd/ /rægd/

In addition, the letter t is silent in the combinations , , , and in several words of French origin, as shown in the word lists below, while in a few common words pronunciations may be found with /t/, or more frequently, without it: often (/ˈɒftən/, /ˈɒfən/), chestnut (/ˈtʃestnʌt/, /ˈtʃesnʌt/), and Christmas (/ˈkrɪstməs/, ˈkrɪsməs/).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

174

Consonants

chasten /ˈtʃeɪsən/ christen /ˈkrɪsən/ fasten /ˈfɑːsən/ glisten /ˈɡlɪsən/ hasten /ˈheɪsən/ listen /ˈlɪsən/ moisten /ˈmɔɪsən/

apostle /əˈpɒsl ̩/ castle /ˈka:sl ̩/ jostle /ˈdʒɒsl ̩/ nestle /ˈnesl ̩/ rustle /ˈrʌsl ̩/ thistle /ˈθɪsl ̩/ whistle /ˈwɪsl ̩/

French loanwords buffet /ˈbʊfeɪ/ gourmet /ˈɡʊəmeɪ/ mortgage /ˈmɔːɡɪdʒ/ nougat /ˈnuːɡɑː/ parquet /ˈpɑːkeɪ/ rapport /ræˈpɔː/ tarot /ˈtærəʊ/ penchant /ˈpɑːnʃɑːn/66

Turning to instances of silent d spellings, they occur in such common words as handkerchief (/ˈhæŋkətʃɪf/) and handsome ˈ(hænsəm), while sandwich and Wednesday have alternative pronunciations with or without /d/, the latter being more common (/ˈsændwɪdʒ/, /ˈsænwɪdʒ/, /ˈwednzdeɪ/, /ˈwenzdeɪ/). In rapid colloquial speech /tj/ and /dj/ are frequently reduced to [tʃ ] and [dʒ], so that no contrast is made between words like due and Jew, the first syllable of Tuesday and the initial sound in choose. Regional and social variants The main variants of English alveolar plosives concern /t/, involving its glottalisation, the amount of aspiration associated with it, as well as its realisation as a brief alveolar tap [ɾ] in intervocalic positions, the latter being a particularly common realisation in high frequency words and expressions (better [ˈbeɾə]) (Collins and Mees 2003: 162; Cruttenden 2014: 178). Unlike American English, in RP there is no tendency to neutralise the /t/ – /d/ contrast in pairs such as coating – coding, rating – raiding, whiter – wider. While the absence of glottal effects is a feature of Conservative RP, the use of glottalised [ʔt] or the replacement of /t/ by a glottal stop [ʔ] is now acceptable in Estuary English67 (get off [ˈɡeʔtɒf] [ˈɡeʔɒf], got it [ˈɡɒʔtɪt] [ˈɡɒʔɪt]). Likewise, the production of [ʔ] preceding syllabic [ɫ ̩] (kettle [ˈkeʔɫ ̩], bottle [ˈboʔɫ̩]) and especially in unaccented intervocalic word-medial positions (later [ˈleɪʔə]) is typical of regional varieties such as Cockney and Glasgow English (see § 5.2.8.2). The amount of aspiration associated with /t/ also varies considerably across accents of English. It is higher in Irish, Indian English and in Welsh- or Gaelicinfluenced English (e.g. Western Highlands) than in RP, but there is generally 66 As we have seen in Section 3.3.1 and you can also see here, the endings and of French loanwords have diphthongal values. 67 Estuary English refers to a variety of British English spoken in the counties adjacent to the estuary of the River Thames, and thus displaying the influence of London regional speech, especially in pronunciation (Crystal 2008: 173–174).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

175

little aspiration in Lancashire, South African and Scottish English. In addition, tap [ɾ] realisations of /t/ in post-accented intervocalic positions, both within words and at close-knit word boundaries, are typical of GA (butter [ˈbʌɾə], put it [ˈpʊɾɪt]), but they may also be heard in South African English and in Southern Irish English, as well as in Cockney as an alternative to [ʔ]. Three other allophonic realisations may also be heard: as post-alveolar approximant [ɹ], in some urban varieties of South Lancashire and West Yorkshire, as retroflex68 [ʈ], in Indian English, whereas in Irish English both /t/ and /d/ are realised as apico-alveolar fricatives [ṱ ḓ] whenever it appears in syllable coda between vowels (bottom [ˈbɑṱəm] and word-finally (but [bʌṱ], wood [wuḓ]) (Hickey 2007: 3128) Comparison with Spanish and advice SSLE should pay particular attention to five issues concerning the articulation of /t/ and /d/ to aim at a native-like pronunciation: (1) adopt an alveolar place of articulation, (2) aspirate [tʰ] in stressed syllable-initial positions, (3) distinguish word final /t/ and /d/, (4) produce /d/ as a plosive in medial and final positions, and (5) not omit /t/ or /d/ in consonant clusters (Finch and Ortiz Lira 1982: 68; Gallardo del Puerto 2005: 196–197; Estebas Vilaplana 2009: 49–62; Mott 2011: 256; Cruttenden 2014: 179). A common pronunciation problem of SSLE is the confusion of the place of articulation /t d/, which is alveolar in RP, the tip of the tongue being in contact with the alveolar ridge, but dental in Spanish, the closure being made between the tip of the tongue and the upper teeth. This is an extremely noticeable error due to the high frequency of these two sounds. In order to master their fluent use in connected speech, SSLE should practice the alveolar stricture, whose recognition and production should not be difficult as it is the same one used to pronounce Spanish /n/. To this effect it can also be helpful to practise homorganic affricated forms of /t d/, [ts dz], (time [ˈtsaɪm], dime [ˈdsaɪm]), avoiding heterorganic dental affrications of the [tθ dð] type (see § 5.2.8.1). Similarly, every effort should be made to produce an aspirated [tʰ] in stressed syllable initial position (tea [tʰiː]), as well as to avoid word-final plosive replacement of /t/ for /d/ keeping in mind that final /t/, but not final /d/ despite being devoiced [d̥], reduces the length of the previous vowel or sonorant (beat [biˑt]) (Dale and Poms 1986; Swan and Smith 2001; Catford 1987; Avery and Ehrlich 1992; Gleason 2012).

68 The label retroflex refers to a place of articulation in which the tip of the tongue is curled back in the direction of the front part of the hard palate (Crystal 2008: 415).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

176

Consonants

Another aspect that causes a strong foreign accent is the tendency that SSLE have to produce a voiced dental fricative [ð] in intervocalic position, as in lado ‘side’ [ˈlaðo], when this is the only environment in which RP /d/ is a fully voiced alveolar stop (ladder [ˈlædər]). Similarly, in word final positions SSLE may mispronounce word-final /d/ as a voiced [ð] or voiceless [θ] fricative because these are the realisations that occur in Spanish (bondad ‘goodness’ [bonðað], [bonðaθ]). This mispronuciation of [ð] for [d] may cause communication problems derived from the loss of the /d/ – /ð/ contrast that distinguishes many English words (breed [ˈbriːd] – breath [ˈbriːð], ride /ˈraɪd/ – writhe /ˈraɪð/). To overcome this pronunciation problem, SSLE are advised to produce English intervocalic and word-final [d] as the initial sound of the word dota ‘supply’, although English /d/ is more retracted. Lastly, the omission of /t/ and /d/ in consonant clusters, especially when they represent grammatical categories such as negation or tense, or when /t/ and /d/ precede homorganic /s/ and /z/ should be avoided as this is one of the factors that most influence the intelligibility and comprehensibility of SSLE. Accordingly, special care should be taken to pronounce final consonants, as for instance in don’t, walked, robbed, Fred’s car, it’s true, what’s this as [dəʊnt], [wɔːkt], [rɒbd], [ˈfredz kɑː], [ɪts ˈtruː] [wɒts ˈðɪs], instead of *[don], *[wɔːk], *[rɒb], *[ˈfrez kɑː], *[ɪs ˈtruː], *[wɒs ˈðɪs]). Further practice AI 4.3 The spellings of RP /t/. Further practice tin /tɪn/ metal /ˈmetl ̩/ bet /bet/

try /traɪ/ twenty /ˈtwenti/ debt /det/

train /treɪn/ knot /nɒt/

It’s twenty to two. /ɪts ˈtwenti tə tuː/ What have you got in your pocket? /ˈwɒt həv ju ˈɡɒt ɪn jə ˈpɒkɪt/ Try internet. /traɪ ˈɪntənet/ The two-twenty-two train tore through the tunnel. /ðə tuː ˈtwenti tuː treɪn ˈtɔː θruː ðə ˈtʌnl ̩/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

AI 4.4 The spellings of RP /d/. Further practice din /dɪn/ medal /ˈmedl ̩/ bed /bed/

drive /draɪv/ added /ˈædɪd/ dinner /ˈdɪnə/

opened /ˈəʊpənd/ middle /ˈmɪdl ̩/

Don’t drink and drive, it’s bad. /dəʊnt drɪŋk ənd draɪv | ɪts bæd/ Dinner will be served downstairs before midnight. /ˈdɪnə wɪl bi sɜːvd ˌdaʊnˈsteəz bɪˈfɔː ˈmɪdnaɪt/ A dozen double damask dinner napkins. /ə ˈdʌzn̩ ˈdʌbl ̩ ˈdæməsk ˈdɪnə ˈnæpkɪnz/ Donald Duck was a Disney character dressed in dark blue and red. /ˈdɒnl ̩d dʌk wəz ə ˈdɪzni ˈkærəktə drest ɪn dɑːk bluː ənd red/ 4.2.1.3 Velar plosives /k ɡ/ IPA Symbols k Lower-case K. ɡ Opentail G. Identification /k/ Fortis voiceless velar plosive. /ɡ/ Lenis voiced velar plosive. Allophones /k/ [kʰ] ‘aspirated’ [k=] ‘unaspirated’ [ʔ] ‘glottalised’ [k̬] ‘voiced’ [k˺] ‘non-audible release’ [kˆ] ‘unreleased’ [kⁿ] ‘nasal release’ [kˡ] ‘lateral release’ [k̟] ‘advanced’ ‘pre-velar’ [k̠] ‘retracted’ ‘post-velar’ [kʷ] ‘labialised’ [tʲ] ‘palatalised’ [x] ‘voiceless velar fricative’

/ɡ/

[ɡ̊] ‘devoiced’ [ɡ˺] ‘non-audible release’ [ɡˆ] ‘unreleased’ [ɡⁿ] ‘nasal release’ [ɡˡ] ‘lateral release’ [ɡ̟] ‘advanced’ ‘pre-velar’ [ɡ̄] ‘retracted’ ‘post-velar’ [ɡʷ] ‘labialised’ [ɡʲ] ‘palatalised’ [ɣ] ‘voiced velar fricative’

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

177

178

Consonants

Description When we pronounce /k/ the two articulators, the back of the tongue and the soft palate, form a complete closure so that the lung airstream is stopped behind it. As the soft palate is in its raised position, there are two closures. In the release stage the two articulators are separated and the lung air goes out of the mouth producing a kind of explosion. The vocal folds are wide apart and do not vibrate. /k/ is produced with more muscular energy than its voiced counterpart, and it shortens the vocalic and sonorant sounds preceding it (leak [liˑk]). /ɡ/ is pronounced in the same way as /k/ but with less muscular effort, and the vocal folds may vibrate for all or part of the compression stage. Aspiration does not take place in the production of this sound, nor does it shorten the length of preceding vocalic or sonorant sounds (league [liːɡ]). See the animations and videos under RP /k/ and /g/ (compared with Spanish /k ɡ/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP /k ɡ/ can occur in all positions in a word. Initially in a stressed syllable /k/ is aspirated [kʰ] (key [kʰiː], accustomed [əˈkʰʌstəmd]), but it loses some of that aspiration [k=] in word-final positions (tick [tɪk=]), and is unaspirated [k=] when preceded by /s/ (scum [ˈsk=ʌm]). Considering voicing, betwen voiced sounds the voiceless velar plosive has a voiced realisation (akin [əˈk̬ɪn]) whereas the voiced velar plosive is fully voiced (anger [ˈæŋɡə], argue [ˈɑːɡjuː], eager [ˈiːɡə], buggy [ˈbʌɡi]), but /g/ loses some of its voicing [ɡ̊] when occurring in initial position (goat [ɡ̊əʊt]) and it is completely devoiced in final position (dog [dɒɡ̊]). /k/ may also be reinforced by glottalisation [ʔ] (actor [ˈæ͜ ʔktə]) when a vowel, nasal or lateral precedes it and it is followed by another consonant or a pause; and in Cockney and other regional (South East) urban dialects /k/ is replaced by the glottal stop word-medially and in consonant clusters (ticket [ˈtɪʔɪt]) (see 5.2.8.2). In addition to the affricated variants (discussed in Section 5.2.8.1), other allophonic variants of both plosives are those with non-audible release [k˺ ɡ˺], if followed by another plosive or by an affricate (locked [ˈlɒk˺t], thick door [ˈθɪk˺ dɔː], rugby [ˈrʌɡ˺bi], big gesture [ˈbɪɡ˺ ˈdʒestʃə]), or those that are unreleased [kˆ ɡˆ], if followed by a homorganic plosive (dark grey [dɑːkˆ ɡreɪ], big game [bɪɡˆ ɡeɪm]). They will be produced with nasal plosion and nasal release [kⁿ ɡⁿ], if followed by a nasal consonant (acknowledge [əkⁿˈnɒlɪdʒ], back movement [ˈbækᵐ ˈmuːvmənt], dogma [ˈdɒɡᵐmə], big novel [ˈbɪɡⁿ ˈnɒvɫ ̩]), and they will have a lateral plosion and lateral release [kˡ ɡˡ] before [l], with the airstream escaping laterally round the point of alveolar closure (clean [kˡliːn], struggle [ˈstrʌɡˡɫ ̩]). In the above examples and elsewhere, notice also that /k/ has a devoicing effect

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

179

on an adjacent sonorant (acknowledge [əkⁿˈn̥ ɒlɪdʒ], clean [kˡl ̥iːn], equal [ˈiːkʷw̥ əl], dockyard [ˈdɒkȷ ̊ɑːd]). Finally, it should be borne in mind that the velar contact is affected by the place of articulation of a subsequent sound (see § 5.2.7). Hence, velar stops will have post-velar realisations [ ]̠ , if followed by a back vowel [uː ʊ ɒ ɔː] or semivowel [w] (cool [k̠uːɫ], good [ɡ̄ ʊd], or pre-velar realisations [ ]̟ , before a front vowel [iː ɪ] or glide consonant [j] (king [k̟ɪŋ̟], geese [ɡ̟i:s]). Similarly, lip position will also vary depending on the articulatory gesture adopted for the adjacent sound. Hence, /k d/ will have allophonic realisations with lip-rounding (cool [kʷuːɫ], quite [kʷwaɪt], goose [ɡʷuːs]) or lip-spreading (kit [kʲɪt], geese [ɡ̟ʲiːs], cure [kʲjʊə], argue [ˈɑːɡʲjuː]) in the vicinity of rounded and spread vowels or semi-vowels, respectively (see § 5.2.3). Spellings Table 22: The spellings of RP /k ɡ/ /k/

/ɡ/

/ + silent key /kiː/, turkey /ˈtɜːki/ make /ˈmeɪk/ gymkhana /dʒɪmˈkɑːnə/, khaki /ˈkɑːki/

hug /hʌɡ/, glad /ɡlæd/

69 cat /kæt/, cold /kəʊld/, act /ækt/

hugged /hʌɡd/, egg /eɡ/

/ / sack /sæk/, blackboard /ˈblækbɔːd/ character /ˈkærəktə/, ache /eik/

/ guide /ɡaɪd/ league /liːɡ/

+ / + occupy /ˈɒkjʊpaɪ/, account /əˈkaʊnt/70 scare /skeər/, discuss /dɪˈskʌs/

+ silent ghastly /ˈɡɑːstli/ ghetto /ˈɡetəʊ/

69 The words Celt and Celtic are pronounced /kelt/ and /ˈkeltik/, but in Scotland they are pronounced /selt/ and /ˈseltɪk/. The names of the Glasgow football (Celtic) and Boston basketball (Celtics) teams also have the latter /s/ pronunciation. 70 Note, however, /ks/ pronunciations of spellings word-medially followed by or as in such words as accent /ˈæksənt/, succeed /səkˈsiːd/ and accident /ˈæksɪdənt/. In addition, other words like flaccid may have two possible pronunciations /ˈflæk.sɪd/ or /ˈflæs.ɪd/, while words borrowed from Italian may have /ʧ/, as in cappuccino /ˌkæpʊˈʧiːnəʊ/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

180

Consonants

+ 71 queen /kwiːn/, quirk /kwɜːk/ liquid /ˈlɪkwɪd/, conquest /ˈkɒŋkwest/ / /ks/ six /sɪks/, axle /æksl ̩/ axe /æks/

/ɡz/ exam /ɪɡˈzæm/, exist /ɪɡˈzɪst/ Alexander /ˌælɪɡˈzɑːndə/

French / Spanish loanwords / queue /kjuː/, liquor /ˈlɪkər/ conquer /ˈkɒŋkər/, quiche /kiːʃ/ mosquito /məˈskiːtəʊ/, quay /kiː/ boutique /buːˈtiːk/, bouquet /bʊˈkeɪ/

lacquer /ˈlækə/, racquet /ˈrækɪt/ + acquire /əˈkwaɪə/, acquaint /əˈkweɪnt/

French loanwords / guitar /ɡɪˈtɑː/ fatigue /fəˈtiːɡ/

Rare cases Akkadian /əˈkeɪdiən/

It should be remembered that spellings with both k/c and g may be silent. The spelling k/c is silent in the and digraphs, as illustrated in the word lists below, as well as in a few other words involving and clusters: indict /ɪnˈdaɪt/, victual(s) /ˈvɪtl ̩/, /ˈvɪtl ̩z/, Connecticut /kəˈnetɪkət/, blackguard /ˈblæɡɑːd/, Cockburn /ˈkəʊbɜːn/, /ˈkəʊbən/. knife /naɪf/ knight /naɪt/ know /nəʊ/ knuckle /ˈnʌkl ̩/

corpuscle /ˈkɔːpʌsl ̩/ muscle /ˈmʌsl ̩/ sceptre /ˈseptə/ scene /si:n/

Turning to , this spelling is silent in the initial and final digraph and in the final cluster .72

71 The digraph is pronounced /kw/ in such words as the aforementioned, but it can also be pronounced /k/ in medial position, especially in words of French or Spanish origin as also illustrated in Table 22. 72 Nevertheless, the g of cluster is pronounced in the following derived words: paradigmatic /ˌpærədɪgˈmatɪk/, phlegmatic /flegˈmætɪk, signal /sɪɡnl ̩/, signatory /ˈsɪɡnətr̩i/, assignation /ˌæsɪɡˈneɪʃn̩/, resignation /ˌrezɪɡˈneɪʃn̩/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

gnaw /nɔː/ gnostic /ˈnɒstɪk/ gnocchi /njɒki/ align /əˈlaɪn/ assign /əˈsaɪn/ benign /bəˈnaɪn campaign /kæmˈpeɪn/ champagne /ʃæmˈpeɪn/ design /dɪˈzaɪn/ physiognomy /ˌfɪzɪˈɒnəmi/ resign /rɪˈzaɪn/ sovereign /ˈsɒvrɪn/

181

diaphragm /ˈdaɪəfræm/ paradigm /ˈpærədaɪm/ phlegm /flem/ syntagm /ˈsɪntæm/

Regional and social variants The main variants of velar plosives concern the degrees of aspiration associated with /k/ (Cruttenden 2014: 182). As in the case of /p t/, more aspiration than in RP is heard in the areas of Celtic influence (Scottish Highlands, Wales and Southern Ireland) and Indian English, whereas lesser aspiration is typical of Lancashire accent. Comparison with Spanish and advice It is essential that SSLE pronounce [kʰ] with aspiration in stressed syllableinitial position key [kʰiː], taking care to reduce the length of the preceding vowel or sonorant when /k/ occurs word finally (leak [liˑk]), in which case aspiration is optional, but generally it is not produced (Dale and Poms 1986; Swan and Smith 2001; Catford 1987; Avery and Ehrlich 1992; Gleason 2012). SSLE should also avoid weakening intervocalic /ɡ/ to a fricative [ɣ], as in lago ‘lake’ [ˈlaɣo], when this is the only environment in which RP /ɡ/ is a fully voiced velar stop (lager [ˈlɑːɡər]), as already noted about the other voiced plosives. In addition, learners should refrain from pronouncing initial [w] as [ɡ] in words such as woman /ˈwʊmən/, which tend to be wrongly pronounced with initial [ɡu-], [ɣu-] or [u-] (Finch and Ortiz Lira 1982: 64, 71; Cruttenden 2014: 182; Estebas Vilaplana 2009: 49–62).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

182

Consonants

Further practice AI 4.5 The spellings of RP /k/. Further practice came /keɪm/ active /ˈæktɪv/ back /bæk/

kicked /kɪkt/ acquire /əˈkwaɪə/ make /ˈmeɪk/

occupy /ˈɒkjʊpaɪ/

I’ve got cake and biscuits. /aɪv ˈɡɒt keɪk ənd ˈbɪskɪts/ I’ll cook the pork at six o’clock. /aɪl kʊk ðə pɔːk ət sɪks əˈklɒk/ That cruel criminal kicked the cat /ðət krʊəl ˈkrɪmɪnl ̩ kɪkt ðə kæt/ Crisp crusts crackle crunchily. /krɪsp krʌsts ˈkrækl ̩ ˈkrʌntʃɪli/

AI 4.6 The spellings of RP /ɡ/. Further practice game /ɡeɪm/ agnostic /æɡˈnɒstɪk/ bag /bæɡ/

gat /ɡəʊt/ hugging /ˈhʌɡɪŋ/ fatigue /fəˈtiːɡ/

guitar /ɡɪˈtɑː/ bigger /ˈbɪɡə/

Give me a bigger book. /ɡɪv miː ə ˈbɪɡə bʊk/ He’s gone to get a green bag. /hiz ɡɒn tə ˈɡet ə ɡriːn bæɡ/ The ground will look greener when the grass has grown. /ðə ɡraʊnd wɪl lʊk ˈɡriːnə wen ðə ɡrɑːs həz ɡrəʊn/ Cows graze in groves on grass which grows in grooves in groves. /kaʊz ɡreɪz ɪn ɡrəʊvz ɒn ɡrɑːs wɪtʃ ɡrəʊz ɪn ɡruːvz ɪn ɡrəʊvz/ 4.2.2 Fricatives RP fricatives /f v θ ð s z ʃ ʒ h/ can be characterised along the following four articulatorily parameters (see also e.g. Gimson 1984: 178–192; Roach 2005: 49– 54; Collins and Mees 2003: 139–148; Collins and Mees 2009: 88–90; Cruttenden 2014: 192–208):

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

183

(1) Manner of articulation: to pronounce a fricative consonant the two articulators used in their production do not come into contact as for the plosives. So, when going out of the mouth through the passage formed by the articulators the lung air produces air turbulence, or a kind of friction or “hissing sound” (Roach 2005: 48). The soft palate is in its raised position thereby closing the nasal passage and releasing the lung air through the oral tract. There is an on/off-glide with respect to the adjacent sound (manifested acoustically by formant transitions), which is more noticeable in the case of vowels. (2) Place of articulation involves five possible points, which is rare in the world’s languages (Maddieson 1984: 43) and is in general a more important distinctive cue than the distinctions between voiced and voiceless pairs of each place of articulation. /f v/ are labio-dental because the two articulators used to produce them are the lower lip, the active articulator, and the upper front teeth, the passive articulator. /s z/ are alveolar, and so the tip of the tongue, the active articulator, rises towards the alveolar ridge, the passive articulator, the airstream escaping along a groove; in contrast /θ ð/ use a flatter configuration, the airstream escaping through the dental slit that is left between the tip of the tongue, the active articulator, and the upper front teeth. /ʃ ʒ/ use a palato-alveolar73 closure that is left between blade of the tongue and the palato-alveolar region, with flat tongue tip and blade and grooving further back, but the latter is not as deep as the one used in /s z/. Finally, /h/ is glottal. This means that the lung air produces friction when it goes out through the glottis and throughout the oral tract, while the articulators are ready in order to produce the following vowel. Phonetically, /h/ can be described as “a voiceless vowel with the quality of the voiced vowel that follows it” (Roach 2005: 60). (3) Position of the vocal folds: the vocal folds do not vibrate when /f θ s ʃ h/ are pronounced, and so these fricative consonants are voiceless or unvoiced. In contrast, when /v ð z ʒ/ are pronounced the vocal folds usually vibrate, but these sounds may lose some of their voicing if they occur initially or finally in the word, and are fully voiced between voiced sounds (Haggard 1978; Docherty 1992). (4) Force of articulation and length: /f θ s ʃ h/ are pronounced with more muscular force or energy than /v ð z ʒ/. So these two subgroups of fricatives are described as fortis or strong, and lenis or weak consonants, respectively. As mentioned before, this feature is quite important because it affects 73 Although in the latest version of the IPA alphabet the fricatives /ʃ ʒ/ and by extension the affricates /tʃ dʒ/ are classified as post-alveolar (see Table 3), in this book the former IPA label palato-alveolar is retained to better reflect the palatalized alveolar nature of these sounds.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

184

Consonants

the length of the preceding vocalic or sonorant sound, which is shortened when followed by a fortis consonant, either within the word or at wordboundaries. For instance, in loath [ləˑʊθ] and roof [ruˑf] the vowel is reduced; while in loathe [ləːʊð] and those [ðəːʊz] the diphthong is fully long because it is followed by lenis [ð] and [z] (see § 2.3.1.4 and 2.3.2.4). Note, however, that although voiceless fricatives shorten the vowels and sonorants that precede them, they are themselves longer than their voiced counterparts (Subtelny et al. 1966; Malécot 1968). Acoustically, as already detailed in Section 2.4.3, fricative consonants are characterised by high-frequency noise (random energy pattern with striations), which is especially strong in voiceless fricatives /f θ s ʃ h/, and even more so in the case of the two sibilants /s ʃ/. Voiced fricatives, on the other hand, show weaker noise and voicing bar, there may being no noise at all when producing the non-sibilant ones /v ð/ (see § 2.4.3). Auditorily, sibilants /s z ʃ ʒ/ are characterised by sound intensity, i.e. they are louder and longer compared to nonsibilants /f v θ ð h/. Under each place of articulation below it will be shown that the main allophones of fricatives concern different degrees of voicing or devoicing (see also § 5.2 and 5.3), as well as the affrication of plosives with [ɸ β s z x ɣ] and the fricativisation of approximants /r l w j/ that are explained in Section § 5.2.8.1. Finally, as regards advice to SSLE, attention should be centred on reproducing with precision the place of articulation required for labio-dental, dental, alveolar and palato-alveolar articulations, and on the reduction of length that fortis fricatives exert on preceding vowel and sonorant sounds (Finch and Ortiz Lira 1982: 59, 60, 68–70; Estebas Vilaplana 2009: 64–81; Cruttenden 2014: 194).

4.2.2.1 Labio-dental fricatives: RP /f v/ vs. PSp /f/ /f v/ IPA Symbols f Lower-case F. v Lower-case V. Identification /f/ Fortis voiceless labio-dental fricative. /v/ Lenis voiced labio-dental fricative.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

Allophones /f/ [f ]̬ ‘voiced’ [f̟] ‘advanced’ [f̠] ‘retracted’ [f ʷ] ‘labialised’ [f ʲ] ‘palatalised’ [v] ‘voiced labio-dental fricative’

185

/v/ [v̥] ‘devoiced’ [v̟] ‘advanced’ [v̠] ‘retracted’ [vʷ] ‘labialised’ [vʲ] ‘palatalised’ [f ] ‘voiceless labio-dental fricative’

Description When we pronounce /f/ the lower lip approximates to the upper front teeth, a small passage is formed, and the soft palate is in its raised position. As a result, when the lung air goes out through that passage it produces friction, a kind of hissing sound. The vocal folds are apart and do not vibrate. /f/ is produced with more muscular energy than its voiced counterpart and it shortens the vocalic sounds that precede it (leaf [liˑf], laugh [lɑˑf], life [laˑɪf]). /v/ is pronounced in the same way as /f/, but, in this case, the vocal folds show different degrees of vocal fold vibration, depending on its phonetic context. /v/ is also produced with less energy than its voiceless counterpart, and it does not shorten the length of preceding vocalic or sonorant sounds (leave [liːv], carve [kɑːv], cave [keːɪv]). See the animations and videos under RP /f/ and /v/ (compared with Spanish /f/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP /f v/ can occur in all word positions. /f/ remains voiceless initially ( fit [fɪt]) and finally (cough [kɒf], but between voiced sounds it acquires some voicing (offer [ˈɒf̬ə]). /v/ is fully voiced betwen voiced sounds (ever [ˈevər], canvas [ˈkænvəs], silver [ˈsɪlvə]), but it loses some of its voicing [v̥] when occurring in initial position (voice [v̥ɔɪs]) and it is completely devoiced in final position (give [ɡɪv̥]). Besides, /v/ may be realised as [f] before a voiceless consonant at closeknit word-boundaries, regularly in have to and more rarely in other sequences (give to, have ten), or it may be elided in rapid, colloquial speech in the unaccented forms of have and of regardless of the following sound (could have played [kəd ə ˈpleɪd], lots of men [ˈlɒts ə ˈmen]). In addition, the labio-dental contact is affected by the place of articulation of a subsequent sound (see § 5.2.7). Hence, labio-dental fricatives will have retracted realisations [ ]̠ , in the vicinity of a bilabial plosive or a back vowel [uː ʊ ɒ ɔː] or semivowel [w] (obvious [ˈɒbv̠ ɪəs] fool [f̠uːɫ]), or advanced realisa-

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

186

Consonants

tions [ ]̟ , before a front vowel [iː ɪ] or glide consonant [j] (feel [f i̟ ːɫ ̩], veal [v̟iːɫ]). Similarly, lip position will also vary depending on the articulatory gesture adopted for the adjacent sound. Hence, /f v/ will have allophonic realisations with lip-rounding ( food [fʷuːd], vodka [ˈvʷɒdkə],) or lip-spreading ( feet [fʲiˑt], vehicle [ˈvʲiˑɪkɫ]) in the vicinity of rounded and spread vowels or semi-vowels, respectively (see § 5.2.3). Spellings Table 23: The spellings of RP /f v/ /f/

/v/

/ prefer /prɪˈfɜːr/, feet /fi:t/ life /laɪf/

/ very /ˈveri/, vulgar /ˈvʌlgə/ give /ɡɪv/, live /lɪv/

/ offer /ˈɒfə/, baffle /ˈbæfl̩/, giraffe /dʒɪˈrɑːf/

navvy /ˈnævi/, savvy /ˈsævi/

, photo /ˈfəʊtəʊ/, phrase /freɪz/ diphthong /ˈdifθɒŋ/, nephew /ˈnefjuː/ sapphire /ˈsæfaɪə/

Stephen /ˈstevn̩/

enough /ɪˈnʌf/, draught /drɑːft/ Rare cases silent + half /hɑːf/ + silent soften /ˈsɒfn̩/

Rare cases

of /əv/

halve /hɑːv/

Regional and social variants /f v/ register no important variants in RP, other than the pronunciation of /f/ as [v] word-initially in South-west England speech (Cruttenden 2014: 197). Comparison with Spanish and advice SLE should avoid pronouncing /v/ as either [b], a voiced bilabial plosive, or [β], a voiced bilabial fricative (Stockwell and Bowen 1965; García Lecumberri 1999, 2000). While in Spanish and spellings correspond to one phoneme, /b/ voiced bilabial plosive (visita ‘visit’ [biˈsita]), in English they represent two different phonemes, voiced bilabial plosive /b/ and voiced labio-dental fricative /v/, respectively. As a result, a common mistake of SSLE is to pronounce such words

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

187

as very /ˈveri/ or visit /ˈvɪzɪt/ with a voiced bilabial plosive (*[ˈbeɾi], *[ˈbɪsɪt]), instead of a voiced labio-dental fricative, which may lead to misunderstandings resulting from the loss of the /b/ – /v/ contrasts that exists in English (very /ˈveri/ – berry /ˈberi/). Care should, therefore, be taken to distinguish /f/, /b/ and /v/ sounds, in fail /feɪl/ – bail /beɪl/ – veil /veɪl/, the last two in particular as they prove to be the most problematic for SSLE. In addition, as discussed in other pairs of consonants, every effort should be made to distinguish word-final [f] and [v̥] by reducing the duration of the vowel or vocalic glide that precedes [f] (leaf /liˑf/ – leave [liːv], safe [seˑɪf] – save [seːɪv̥]) (Finch and Ortiz Lira 1982: 64, 71; Dale and Poms 1986; Swan and Smith 2001; Catford 1987; Avery and Ehrlich 1992; Cruttenden 2014: 182; Estebas Vilaplana 2009: 80–82; Gleason 2012). Further practice AI 4.7 The spellings of RP /f/. Further practice fat /fæt/ rifle /ˈraɪfəl/ leaf /liːf/

family /ˈfæməli/ offer /ˈɒfə/ off /ɒf/

photo /ˈfəʊtəʊ/ giraffe /dʒɪˈrɑːf/

We finally voted on Friday at five fifteen. /wi ˈfaɪnəli ˈvəʊtɪd ɒn ˈfraɪdi ət faɪv ˌfɪfˈtiːn/ My nephew laughs at funny jokes about fishing. /maɪ ˈnevjuː lɑːfs ət ˈfʌni dʒəʊks əˈbaʊt ˈfɪʃɪŋ/ I live with five friends in a fifth-floor furnished flat. /ˈaɪ ˈlɪv wɪð faɪv frendz ɪn ə fɪfθ flɔː ˈfɜːnɪʃt flæt/ Four furious friends fought for the phone. /fɔː ˈfjʊərɪəs frendz ˈfɔːt fə ðə fəʊn/ Fred fed Ted bread, and Ted fed Fred bread. /fred fed ted bred | ənd ted fed fred bred/

AI 4.8 The spellings of RP /v/. Further practice vat /væt/ rival /ˈraɪvəl/ leave /liːv/

very /ˈveri/ travel /ˈtrævl ̩/ give /ɡɪv/

view /vjuː/ seven /ˈsevn̩/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

188

Consonants

Various student groups travelled to Venice in the vacation. /ˈveərɪəs ˈstjuːdnt ɡruːps ˈtrævl ̩d tə ˈvenɪs ɪn ðə vəˈkeɪʃn̩/ A very wide variety of people came. /ə ˈveri waɪd vəˈraɪəti əv ˈpiːpl ̩ keɪm/ I love the nice view over the village. /| ˈaɪ lʌv ðə naɪs vjuː ˈəʊvə ðə ˈvɪlɪdʒ/ That vet looks after virtually all the livestock in the valley. /ðət vet lʊks ˈɑːftə ˈvɜːtʃʊəli ɔːl ðə ˈlaɪvstɒk ɪn ðə ˈvæli/. Vincent vowed vengeance very vehemently. /ˈvɪnsn̩t vaʊd ˈvendʒəns ˈveri ˈviːəməntli/ Seventy seven benevolent elephants. /ˈsevn̩ti ˈsevn̩ bəˈnevələnt ˈelɪfənts/ 4.2.2.2 Dental fricatives: RP /θ ð/ vs. PSp /θ/ IPA Symbols θ Theta. ð Eth. Identification /θ/ Fortis voiceless dental fricative. /ð/ Lenis voiced dental fricative. Allophones /θ/ [θ̬] ‘voiced’ [θʷ] ‘labialised’ [θʲ] ‘palatalised’ [ð] ‘voiced dental fricative’ [f ] ‘voiceless labio-dental fricative’

/ð/ [ð̥] ‘devoiced’ [ðʷ] ‘labialised’ [ðʲ] ‘palatalised’ [θ] ‘voiceless dental fricative’ [v] ‘voiced labio-dental fricative’ [z] ‘voiced alveolar fricative’ [l] ‘alveolar lateral approximant’

Description To pronounce /θ/ the soft palate is in its raised position, and the tip of the tongue makes a light contact with the edge and inner surface of the upper incisors so that a small passage is formed, through which the outgoing airflow is released producing a slight friction. The vocal folds do not vibrate, but are wide apart. /θ/ is produced with more muscular energy than its counterpart, and it shortens the vocalic sounds that precede it (birth [bɜˑθ], wreath [ɹiˑθ],

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

189

oath [əˑʊθ]). /ð/, on the other hand, is pronounced in the same way as /θ/, but with less energy and with the vocal folds showing different degrees of vibration depending on its context of occurrence. It does not shorten the length of preceding vocalic or sonorant sounds (wreathe [ɹiːð]). For the articulation of both /θ ð/ the tongue is quite flat, and as a result, the passage through which the lung air is released has “the nature of a slit rather than a groove” (Gimson 1994: 168; Cruttenden 2014: 199; Collins and Mees 2003: 141). For this reason these two dental consonants are technically known as slit fricatives (Stone 1990; Stone et al. 1992) (see § 2.3.2.3). See the animations and videos under RP /θ/ and /ð/ (compared with Spanish /θ/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP /θ ð/ can occur in all word positions. /θ/ remains voiceless initially (thick [θɪk]) and finally (breath [breθ], but between voiced sounds it gains some voicing (ethics [ˈeθ̬ɪks]). /ð/ is fully voiced betwen voiced sounds (leather [ˈleðər], although [ɔːɫˈðəʊ]), but it loses some of its voicing [ð̥] when occurring in initial position (there [ð̥eə]) and it is completely devoiced in final position (clothe [kləʊð̥], with [wɪð̥]). Other allophonic realisations with lip-rounding (thought [ˈθʷɔˑt], soothe [suːðʷ]) or lip-spreading (thin [θʲɪn], this [ðʲɪs]) will also emerge in the vicinity of rounded [uː ʊ ɒ ɔː w] and spread [iː ɪ j] vowels or semi-vowels, respectively (see § 5.2.3). In addition, /ð/ may be pronounced as [z], with an alveolar place of articulation, at close-knit word boundaries in rapid speech if preceded by /z/ (Is there some? [ˈɪz zeə ˈsʌm]). Both dental fricatives may also be elided in consonant clusters preceding /s z/ (twelfths [twelfs], clothes [kləʊz]) to ease difficulties of articulation (see § 5.4). Spellings Table 24: The spellings of RP /θ ð/ /θ/

/ð/

thin /θɪn/, through /θruː/ author /ˈɔːθə/ cathedral /kəˈθiːdrəl/ north /nɔːθ/, worth /ˈwɜːθ/

then (weak) /ðen/ rather (strong) /ˈrɑːðə/ worthy /wɜːði/

breathe /briːð/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

190

Consonants

There are some pronunciation hints that may be helpful when deciding whether the spelling is pronounced as /θ/ or /ð/. In initial position, the digraph th is pronounced /ð/ in most function words and adverbs (the, this, theirs, than, there, thus, then) with the exception of through /θruː/; but in most other words it is pronounced /θ/. In medial position, it tends to be pronounced /ð/ (bother, brother, either, gather, other, rather), but there are a few exceptions which have /θ/ such as panther /ˈpænθə/, ether /ˈiːθə/ and compounds of thing /θɪŋ/ (everything, something, etc.). In final positions, the th spelling is usually pronounced /θ/, with two exceptions, with /wɪð/ and smooth /smuːð/; whereas in some other cases both [ð] and [θ] pronunciations can be heard (booth [buːð], [buˑθ], bequeath [bɪˈkwiːð], [bɪˈkwiˑθ]). Generally, however, [ð] – [θ] pronunciation differences usually serve to mark noun [θ] – adjective or verb [ð] contrasts, or otherwise they identify noun-verb cognates, spelt with final th and the, respectively, as illustrated in the following examples: Noun North /nɔːθ/ South /saʊθ/ worth /wɜːθ/

Adjective northern /ˈnɔːðən/ southern /ˈsʌðən/ worthy /ˈwɜːði/

Noun bath /bɑːθ/ breath /breθ/ sheath /ʃiːθ/ sooth /suːθ/ wreath /riːθ/

Verb bathe /beɪð/ breathe /briːð/ sheathe /ʃiːð/ soothe /suːð/ wreathe /riːð/

Regional and social variants Dental fricatives have no significant variants in RP. In London speech /ð/ may be pronounced as labio-dental [v] mother [ˈmʌvə], or with such alveolar articulations as [l] (all the [ɔːɫ lə]) and [n] (in the [ɪn nə]), when /ð/ occurs initially in weak forms after /t/ and /n/ at close-knit word-boundaries (Kerswill 2003). Further allophonic realisations may be heard in Irish English when both dental fricatives are realised as dentalised plosives [t ̪ d̪] (thank [t ̪ænk]), although word-finally they can be pronounced as fricatives (path [pʰæθ]) (Wells 1982: 428–431; Cruttenden 2014: 199; Hickey 1984: 240; Hughes 2005: 116). Comparison with Spanish and advice The [θ] sound produced in most of Spain (caza ‘hunting’ [ˈkaθa]) makes an acceptable equivalent of RP /θ/, even though the Spanish sound is fully interdental and is produced with considerably more friction. [ð], on the other hand, does not have phonemic status in Spanish, but is used as an allophone of [d] in all positions except before a pause and after a

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

191

nasal (cada ‘each’ [kaða] (Navarro Tomás 1966 [1946]; Quilis and Fernández 1982). A common mistake of SSLE is therefore to pronounce [d] in the place of [ð] in words such as the, then, their, them, there, etc. This error should be avoided because, although it may not be critical for understanding, it may hinder comprehension. Finally, it should be kept in mind that the main cue to distinguish between word-final [θ] and [ð̥] is that the vowels that precede the voiceless phoneme (teeth [tiˑθ]) are shorter than those occurring before the voiced one (teethe [tiːð̥]) (Finch and Ortiz Lira 1982: 69; Gallardo del Puerto 2005: 201; Estebas Vilaplana 2009: 77–79; Sedláčková 2010: 59; Mott 2011: 257; Gleason 2012; Cruttenden 2014: 173). Further practice AI 4.9 The spellings of RP /θ/. Further practice thin /θɪn/ healthy /ˈhelθi/ tooth /tuːθ/

through /θruː/ twenty-third /ˈtwenti θɜːd/ bath /bɑːθ/

think /θɪŋk/ cloth /klɒθ/

King Arthur sat on his throne. /kɪŋ ˈɑːθə sæt ɒn ɪz θrəʊn/ I think this thing is too thin. /ˈaɪ θɪŋk ðɪs ˈθɪŋ z tuː θɪn/ The bath is only thirty-three inches wide. /ðə bɑːθ s ˈəʊnli ˈθɜːti θriː ˈɪntʃɪz waɪd/ The thirty-three thieves thought that they thrilled the throne throughout Thursday. /ðə ˈθɜːti θriː θiːvz ˈθɔːt ðət ˈðeɪ θrɪld ðə θrəʊn θruːˈaʊt ˈθɜːzdi/

AI 4.10 The spellings of RP /ð/. Further practice this /ðɪs/ other /ˈʌðə/ smooth /smuːð/

then /ðen/ teething /ˈtiːðɪŋ/ clothe /kləʊð/

rather /ˈrɑːðə/ there /ðeə/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

192

Consonants

I breathe rather fast when I’m bathing. /ˈaɪ briːð ˈrɑːðə fɑːst wen aɪm ˈbeɪðɪŋ/ I’ll do this work here then, and that work there. /aɪl də ðɪs ˈwɜːk hɪə ðen | ənd ðət ˈwɜːk ðeə/ It was the coldest day of the year. /ɪt wəz ðə ˈkəʊldɪst deɪ əv ðə jɪə/ Don’t bother to gather all the leaves and heather. /dəʊnt ˈbɒðə tə ˈɡæðər ɔːl ðə liːvz ənd ˈheðə/ There those thousand thinkers were thinking: how did the other three thieves go through? /ðeə ðəʊz ˈθaʊzn̩d ˈθɪŋkəz wə ˈθɪŋkɪŋ ˈhaʊ dɪd ði ˈʌðə θriː θiːvz ɡəʊ θruː/ 4.2.2.3 Alveolar fricatives: RP /s z/ vs. PSp /s/ IPA Symbols s Lower-case S. z Lower-case Z. Identification /s/ Fortis voiceless alveolar fricative. /z/ Lenis voiced alveolar fricative. Allophones /s/ [s̬] ‘voiced’ [s̠] ‘retracted’ [s]̟ ‘advanced’ [sʷ] ‘labialised’ [sʲ] ‘palatalised’ [z] ‘voiced alveolar fricative’

/z/ [z̥] ‘devoiced’ [z̠] ‘retracted’ [z̟] ‘advanced’ [zʷ] ‘labialised’ [zʲ] ‘palatalised’

Description To pronounce /s/, the blade (and tip) of the tongue (Bladon and Nolan 1977) are quite close to the alveolar ridge and the side rims of the tongue are placed against the upper side teeth. The soft palate is in its raised position so that the outgoing lung air is released through an oral passage that has the shape of a narrow groove in the centre of the tongue − as opposed to the slit of dental fricatives −, producing friction (Gimson 1994: 170; Collins and Mees 2009: 89; Cruttenden 2014: 202). For this reason /s z/ are technically known as grooved fricatives (Stone 1990; Stone et al. 1992) (see § 2.3.2.3.). /s/ shortens the long

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

193

vocalic sounds that precede it (cease [siˑs], grass [ɡɹɑˑs]), and is pronounced with more muscular energy than its voiced counterpart. /z/, in turn, is pronounced in the same way as /s/, but has potential voice and is produced with less energy than its voiceless counterpart /s/, without affecting the length of previous vocalic and sonorant sounds (seize [siːz]). See the animations and videos under RP /s/ and /z/ (compared with Spanish /s/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP /s z/ can occur in all word positions. /s/ remains voiceless initially (sin [sɪn]) and finally (cats [kæts], but between voiced sounds it gains some voicing (essay [eˈs̬eɪ]). /z/ is fully voiced betwen voiced sounds (easy [ˈiːzi], husband [ˈhʌzbənd]), but it loses some of its voicing [z̥] when occurring in initial position (zinc [z̥ɪŋk]) and it is completely devoiced in final position (does [dʌz̥]). In addition, /s z/ will have retracted realisations [ ]̠ before /r/ and in the vicinity of a back vowel [uː ʊ ɒ ɔː] or semivowel [w] (horse-riding [hɔː s̠ˈɹaɪdɪŋ] newsreel [ˈnjuːz̠ɹiːl]), or advanced realisations [ ]̟ , before a front vowel [iː ɪ] or glide consonant [j] (seal [s̟iːɫ], zero [ˈz̟ɪərəʊ]) (see § 5.2.7). In the same contexts, i.e. in the vicinity of rounded [uː ʊ ɒ ɔː w] and spread [iː ɪ j] vowels or semivowels, the realisations of /s z/ will involve lip-rounding (soon [sʷuːn], zoo [zʷuː]) and lip-spreading (see [sʲiː], zeal [zʲiːɫ]) respectively (see § 5.2.3). Spellings Table 25: The spellings of RP /s z/ /s/

/z/

/ / cats /kæts/ a cat’s tail /ə ˈkæts teɪl/ cats’ tails /kæts teɪlz/

/ / boys /ˈbɔɪz/ a boy’s brain /ə ˌbɔɪz breɪn/ boys’ brains /ˈbɔɪz breɪnz/

74 false /fɔːls/, goose /ɡuːs/

/ / 75 please /pliːz/, goes /ɡəʊz/ Chinese /tʃaɪˈniːz/ Portuguese /ˌpɔːtʃʊˈɡiːz/

miss /mɪs/, missile /ˈmɪsaɪl/

76 dessert /dɪˈzɜːt/, scissors /ˈsɪzəz/

74 As you can see in Table 25, word final may be pronounced either /s/ (else /els/, cease /siːs/) or /z/ (rise /raɪz/, raise /reɪz/). 75 The adjective and noun suffix -ese is always pronounced /iːz/. 76 is pronounced /z/ in approximately five words (e.g. possess /pəˈzes/, dissolve /dɪˈzɒlv/, hussar /hʊˈzɑː/ and the two other words included in Table 25), its NP being /s/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

194

Consonants

/ 77 centre /ˈsentə/, council /ˈkaʊnsl ̩/ romance /rəʊˈmæns/, sauce /sɔːs/ once /wʌns/ rice /raɪs/

/ 78 zoo /zuː/ breeze /briːz/, maze /meɪz/ seize /siːz/

/ ascend /əˈsend/ acquiesce /ˌækwiˈes/

jazz /dʒæz/ puzzle /ˈpʌzl ̩/

blitz /blɪts/

xylophone /ˈzaɪləfəʊn/

/ /ks/ six /sɪks/, axle /æksl ̩/ axe /æks/

/gz/ exam /ɪɡˈzæm/, exist /ɪɡˈzɪst/ Alexander /ˌælɪɡˈzɑːndə/

Before moving on to the next heading, let us focus on some pronunciation hints. The noun suffix -ism is always pronounced [ɪzəm] or [ɪzm̩ ], as in athleticism ([æθˈletɪˌsɪzəm], [æθˈletɪˌsɪzm̩ ]), capitalism ([ˈkæpɪtəlɪzəm], [ˈkæpɪtəlɪzm̩ ]), Marxism ([ˈmɑːksɪzəm], [ˈmɑːksɪzm̩ ]), racism ([ˈreɪsɪzəm], [ˈreɪsɪzm̩ ]). In addition, there exist /s/-/z/ alternations in pronunciation, which usually trigger verb-noun or adjective contrasts that tend, but need not be signalled orthographically with / spellings respectively, as can be seen in the following examples: Verb abuse /əˈbjuːz/ advise /ədˈvaɪz/ close /kləʊz/ devise /dɪˈvaɪz/ diffuse /dɪˈfjuːz/ excuse /ɪkˈskjuːz/ house /ˈhaʊz/ use /ˈjuːz/79

Noun or Adjective /əˈbjuːs/ advice /ədˈvaɪs/ /kləʊs/ device /dɪˈvaɪs/ /dɪˈfjuːs/ /ɪkˈskjuːs/ /ˈhaʊs/ /ˈjuːs/

Nevertheless, there exist some exceptions to the spelling-pronounciation correlations just described. For instance, although the se spellings of license and practise serve to distinguish these verbs from the nouns licence and practice, both verbs and nouns have the same pronunciation with a final /s/ (/ˈlaɪsn̩s/, /ˈpræktɪs/). Likewise, decrease, increase, promise and release are also pronounced with final /s/ regardless of their use as verbs or nouns (/dɪˈkriːs/, /ɪnˈkriːs/, /ˈprɒmɪs/, /rɪˈliːs/). In contrast, fuse and surprise are always pronounced with final /z/ (/fjuːz/, /səˈpraɪz/). 77 Word final is always pronounced /s/, never /z/. 78 Word final is always pronounced /z/, never /s/. 79 Note, however, that in the habitual auxiliary used to it is always /ju:st/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

195

Turning to the endings corresponding to regular plurals, possessive nouns and the third person singular simple present tense of regular verbs, they may have three pronunciations, [s], [z] and [ɪz] or [əz], which are predictable from the last sound of the base to which they are attached. If the noun or verb ends in a sibilant consonant /s z ʃ ʒ tʃ dʒ/, then the ending is realised as [ɪz] or [əz]: church /tʃɜːtʃ/ garage /ˈɡærɑːʒ/ George /dʒɔːdʒ/ lass /læs/ lash /læʃ/ noise /nɔɪz/ clash /klæʃ/

churches /ˈtʃɜːtʃɪz/ garages /ˈɡærɑːʒɪz/ George’s /ˈdʒɔːdʒəz/ lasses /ˈlæsɪz/ lashes /ˈlæʃɪz/ noises /ˈnɔɪzɪz/ clashes /klæʃɪz/

When the noun or verb ends in a vowel or a voiced non-sibilant consonant, the ending is pronounced [z]: Bob /bɒb/ eagle /ˈiːɡl ̩/ fee /fiː/ say /seɪ/ prism /ˈprɪzəm/

Bob’s /bɒbz/ eagles /ˈiːɡl ̩z/ fees /fi:z/ pays /peɪz/ prisms /ˈprɪzəmz/

If the noun or verb ends in a voiceless non-sibilant consonant, the ending is realised as [s]: belief /bɪˈliːf/ cough /kɒf/ opposite /ˈɒpəzɪt/ sixth /sɪksθ/ thank /θæŋk/

beliefs /bɪˈliːfs/ coughs /kɒfs/ opposites /ˈɒpəzɪts/ sixths /sɪksθs / thanks /θæŋks/

Finally, attention should be paid to some vestiges of the Old English pronunciation rule according to which fricative consonants should be voiced in intervocalic position. As a result, some singular nouns ending in /ð/ exhibit /ðz/ plural patterns, while the rest either admit plurals with both [θs] and [ðz] pronunciations or otherwise only have /θs/ plurals, as shown below:80 80 In some of these nouns alternative pronunciations may also be heard but they fall outside mainstream RP (see fn 9). Another instance of the fossilization of this Old English phonotactic rule explains the pronunciation of house-houses /haʊs/-/ˈhaʊzɪz/ as no other singular noun ending in /s/ forms its plural in this way in modern English.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

196

Consonants

plurals with /ðz/ mouth-mouths /maʊθ/-/maʊðz/

plurals with [ðz] or [θs] truth-truths /truːθ/-[truːðz], [truːθs]

plurals with /θs/ death-deaths /deθ/-/deθs/

path-paths /pɑːθ/-/pɑːðz/

bath-baths /bɑːθ/-[bɑːðz], [bɑːθs]

month-months /mʌnθ/-/mʌnθs/

youth-youths /juːθ/-/juːðz/

lath-laths /lɑːθ/-[lɑːθs], [lɑːðz]

moth-moths /mɒθ/-/mɒθs/

cloth-cloths /klɒθs/-[klɒθs], [klɒðz]

myth-myths /mɪθ/-/mɪθs/

Regional and social variants In regional varieties of RP, while /nz/ and /ndz/ clusters are kept distinct by most speakers (wins-winds, fines-finds), the pronunciation of the cluster /ns/ as /nts/ with an epenthetic (or additional) alveolar plosive is widespread, thereby neutralizing to /nts/ pronunciations the distinctions between such words as assistance-assistants, mince-mints, plans-plants (see § 5.5). Similarly, in clusters consisting of s + /m/ or /ŋ/ an epenthetic plosive that is homorganic with the nasal may also be heard (Samson [ˈsæmsən] [ˈsæmpsən], Kingston [ˈkɪŋstən] [ˈkɪŋkstən]). In addition, in South-West England /s/ may be replaced by /z/ in word-initial position (serve [z̥ɜːv]) (Cruttenden 2014: 202–203). Comparison with Spanish and advice Although the RP /s/ and the Andalusian and Latin American variety of [s] may be regarded as acceptable equivalents, both being made with the blade of the tongue raised towards the alveolar region, the type of [s] produced in the rest of Spain is apical: the stricture is made between the tongue tip and the alveolar ridge. Since apical [s] has a slight [ ʃ ] quality, the [s]-[ ʃ ] contrast should be practised. Besides, it should be noted that [s] is usually dropped and replaced by [h] in pre-consonantal and final positions in Latin-American Spanish (Argentina, Chile, Central America) and in Andalusian and Canary Spanish, which does not occur in RP. Speakers of Spanish should also avoid the addition of a vowel before any of the clusters beginning with /s/ (scale / skeɪl/). A good practice in this case is to pronounce a long [sss] before the adjacent consonant. The potential problems of RP /z/ should not concern its articulation because in PSp and in some varieties of Latin-American Spanish [z] occurs as an allophone of /s/ before voiced consonants (e.g. mismo ‘same’ [ˈmizmo]), and it is the sound found in such Catalan words as casa ‘house’ [ˈkaza]. The difficulty rather resides in that in English /s z/ have phonemic status and show a much freer occurrence than the Spanish equivalent sounds, of which only /s/ has

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

197

phonemic nature. As a result SSLE may find it difficult to produce and perceive the differences between /s/ and /z/, and misunderstandings may occur since such English words as Sue /suː/ – /sjuː/ and zoo /zuː/ may be pronounced in the same way ([suː]). To grasp the /s/ – /z/ distinction is of utter importance because these two sounds are very frequent in English. They are used together with /ɪz/ (or /əz/) to mark plurals (legs /leɡz/), the third person singular of verbs (eats /iːts/) and the possessive case (George’s /ˈdʒɔːdʒɪz/), as well as in auxiliary verbs (doesn’t /ˈdʌznt/ and isn’t /ˈɪznt/) and many grammatical words, among other contexts. SSLE usually pronounce spellings as [s] (*[dʌsnt] and *[ɪsnt]) or *[is] (Sedláčková 2010: 59; Estebas Vilaplana 2009: 107–109), which are close to the devoiced allophones [z̥] [ɪz̥] of the target pronunciations. However, SSLE should try to attain [z] and [z̥] (by weakening), lengthening at the same time the sound preceding it (bags [bæːɡːz̥]). In contrast, when pronouncing [s] the previous vocalic and sonorant sounds should be shortened (backs /bæˑks/). In addition, the differences with the other sibilants or fricatives should also be practised, especially if they occur successively within the same word or phrase (e.g. decision /dɪˈsɪʒn̩/, scissors /ˈsɪzəz/, thousand /ˈθaʊzn̩d/, enthusiasm /ɪnˈθjuːzɪæzəm/, etc.) (Finch and Ortiz Lira 1982: 69; Estebas Vilaplana 2009: 73– 76; Cruttenden 2014: 203). Further practice AI 4.11 The spellings of RP /s/. Further practice peace /piːs/ mercy /ˈmɜːsi/ course /kɔːs/

centre /ˈsentə/ castle /ˈkɑːsl ̩/ romance /rəʊˈmæns/

silly /ˈsɪli/ fleece /fliːs/

Some say Suzan is silly, but Alice is sillier. /səm ˈseɪ ˈsuːzn̩ z ˈsɪli | bət ˈælɪs ɪz ˈsɪlɪə/ Ice in juice is nice. /aɪs ɪn dʒuːs ɪz naɪs/ Six terrorists assassinated the prince’s son. /sɪks ˈterərɪsts əˈsæsɪneɪtɪd ðə ˈprɪnsəz sʌn/ Six sick hicks nick six slick bricks with picks and sticks. /sɪks sɪk hɪks nɪk sɪks slɪk brɪks wɪð pɪks ənd stɪks/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

198

Consonants

AI 4.12 The spellings of RP /z/. Further practice peas /piːz/ dizzy /ˈdɪzi/ freeze /friːz/

zebra /ˈzebrə/ crazy /ˈkreɪzi/

fleas /fliːz/ eyes /aɪz/

Zebras are like horses but striped. /ˈzebrəz ə ˈlaɪk ˈhɔːsɪz bət straɪpt/ Zoos are crazy places! /zuːz ə ˈkreɪzi ˈpleɪsɪz/ Rosa plays the xylophone outside. /ˈrozə pleɪz ðə ˈzaɪləfəʊn ˌaʊtˈsaɪd/ A noisy noise annoys an oyster. /ə ˈnɔɪzi nɔɪz əˈnɔɪz ən ˈɔɪstə/ Denise sees the fleece, Denise sees the fleas. /dəˈniːz ˈsiːz ðə fliːs | dəˈniːz ˈsiːz ðə fliːz/ At least Denise could sneeze and feed and freeze the fleas. /ət liːst dəˈniːz kəd sniːz ənd fiːd ənd friːz ðə fliːz/ 4.2.2.4 RP palato-alveolar fricatives /ʃ ʒ/ IPA Symbols ʃ Esh. ʒ Ezh. Identification /ʃ/ Fortis voiceless palato-alveolar fricative. /ʒ/ Lenis voiceless palato-alveolar fricative. Allophones /ʃ/ [ʃ ]̬ ‘voiced’ [ʃʷ] ‘labialised’ [ʃʲ] ‘palatalised’ [ʒ] ‘voiced palato-alveolar fricative’

/ʒ/ [ʒ̊] ‘devoiced’ [ʒʷ] ‘labialised’ [ʒʲ] ‘palatalised’ [ʃ] ‘voiceles palato-alveolar fricative’ [dʒ] ‘voiced palato-alveolar affricate’

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

199

Description The nasal resonator being shut off by raising the soft palate, the tip and blade of the tongue makes a light contact with the alveolar ridge, and the front of the tongue is raised in the direction of the hard palate while the side rims of the tongue are in contact with the upper side teeth. The pulmonic air stream escapes causing friction through a more diffuse and extensive area (compared with /s z/) between the tongue and the roof of the mouth. As a fortis voiceless consonant produced without vocal fold vibration, /ʃ/ shortens the vocalic and sonorant sounds that precede it (moustache [məˈstɑˑʃ ], douche [duˑʃ ]). /ʒ/, in turn, is pronounced with less energy, and as a lenis voiced consonant, it may show different degrees of vocal fold vibration depending on the context, but it does not reduce the length of preceding sounds (rouge [ruːʒ], garage [ˈɡærɑːʒ]). See the animations and videos under RP /ʃ/ and /ʒ/ in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations /ʃ ʒ/ may occur in all contexts, albeit the particularly weak functional load of /ʒ/. It shows a more restrictive distribution than its voiceless counterpart because it only exists word-finally in French loanwords, and it is rarely found wordinitially and never occurs in word-initial consonantal clusters. Besides, /ʒ/ may be pronounced as [dʒ] word-finally in fully anglicised forms, and it may alternate with [ ʃ ] word-medially unless these different pronunciations alter the meanings of the words (measure /ˈmeʒə/ vs. mesher /ˈmeʃə/). /ʃ/ remains voiceless initially (shed [ ʃed]) and finally ( finish [ˈfɪnɪʃ ], but between voiced sounds it gains some voicing (mission [ˈmɪʃn̩̬ ], rashly [ˈræʃli]). ̬ /ʒ/ is completely voiced when it occurs between voiced sounds (measure [ˈmeʒə]), but in all the other cases it is devoiced (genre [ˈʒ̊ɑːnrə], gendarme [ˈʒ̊ɒndɑːm], rouge [ɹuːʒ̊]). As regards lip position, the realisations of /ʃ ʒ/ will involve lip-rounding (sugar [ˈʃʷʊɡə], rouge [ɹuːʒʷ]) and lip-spreading (sheer [ ʃ ʲɪə], gigolo [ˈʒʲɪɡələʊ]) in the vicinity of rounded [uː ʊ ɒ ɔː w] and spread [iː ɪ j] vowels or semi-vowels, respectively (see § 5.2.3).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

200

Consonants

Spellings Table 26: The spellings of RP /ʃ ʒ/ /ʃ/

/ʒ/

ship /ʃɪp/, brush /brʌʃ/ publisher /ˈpʌblɪʃə/

vision /ˈvɪʒn̩/, fusion /ˈfjuːʒn̩/

, expansion /ɪkˈspænʃn̩/ profession /prəˈfeʃn̩/

+ / + casual /ˈkæʒʊəl/ seizure /ˈsiːʒə/

/ + sugar /ˈʃʊɡə/, pressure /ˈpreʃə/ assure /əˈʃɔː/ / + nation /ˈneɪʃn̩/, function /ˈfʌŋkʃn̩/ negotiate /nɪˈɡəʊʃɪeɪt/ , delicious /dɪˈlɪʃəs/ ocean /ˈəʊʃn̩/ / + / conscious /ˈkɒnʃəs/ fascist /ˈfæʃɪst/ crescendo /krɪˈʃendəʊ/ French / German Loanwords / machine /məˈʃiːn/ quiche /kiːʃ/ / fuchsia /ˈfjuːʃə/, schedule /ˈʃedjuːl/ schmuck /ʃˈmʌk/, schmaltz /ʃmɔːlts/

French loanwords / genre /ˈʒɑːnrə/, regime /reɪˈʒiːm/ massage /ˈmæsɑːʒ/, beige /beɪʒ/

bijou /ˈbiːʒuː/ déjà vu /ˌdeɪʒə ˈvuː/ je ne sais quoi /ˌʒə nə seɪ ˈkwɑː/

Regional and social variants /ʃ ʒ/ have no significant regional or social variations. The most remarkable one is that, while Conservative RP speakers prefer to use [s] or [z] + /i/ or /j/ sequences, alternative pronunciations may be heard in two contexts: (1) before /uː/ or /ʊə/, where there are fluctuations between /ʃ ʒ/ pronunciations and those with [s] or [z] + /j/ (tissue [ˈtɪʃuː] [ˈtɪsjuː], usual [ˈjuːʒʊəl] [ˈjuːʒwəl]; (2) before other vowels, where a similar variation exists between /ʃ/ and /si/ or /sj/ (ratio [ˈɹeɪʃɪəʊ]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

201

[ˈɹeɪsjəʊ], negotiate [nɪˈɡəʊʃieɪt] [nɪˈɡəʊsjeɪt]). In certain words before /ə/, however, pronunciations with the sequences [s] or [z] + /i/ or /j/ are preferred, especially in comparatives (Parisian [pəˈrɪzɪən], easier [ˈiːzɪə]) (Cruttenden 2014: 205). Comparison with Spanish and advice As Spanish lacks /ʃ ʒ/ in its phonemic inventory, these two phonemes tend to be substituted by other sounds (Dale and Poms 1986; Swan and Smith 2001; Catford 1987; Avery and Ehrlich 1992; Estebas Vilaplana 2009: 64–66; Gleason 2012; Cruttenden 2014: 206). /ʃ/ tends to be pronounced as [s] or [tʃ ], and so such words as show /ʃəʊ/ or Washington /ˈwɒʃɪŋtən/ may be wrongly pronounced as *[tʃou] and *[ˈwosɪntən]. To avoid such mispronunciations, learners could try to produce the /ʃ/ sound corresponding to the spelling in Galician and Catalan (Caixa) or to in the Andalusian accent spoken in Cádiz. Similarly, instead of /ʒ/, speakers of Spanish commonly use [ ʃ ] or even [s], and so instead of division /dɪˈvɪʒn̩/, pleasure /ˈpleʒə/ or leisure /ˈleʒə/, such pronunciations as *[diˈvɪsion], *[pleʃǝ] and *[lesur] may be heard. SSLE of Basque origin may have no difficulty in producing this sound as it is used in some Vizcayan dialects (García Lecumberri and Elorduy 1994). Further practice AI 4.13 The spellings of RP /ʃ/. Further practice shin /ʃɪn/ nation /ˈneɪʃn̩/ wash /wɒʃ/

shoot /ʃuːt/ machine /məˈʃiːn/ delicious /dɪˈlɪʃəs/

cash /kæʃ/ push /pʊʃ/ station /ˈsteɪʃn̩/

I wish the sun would shine like in Asia. /ˈaɪ wɪʃ ðə sʌn wʊd ʃaɪn ˈlaɪk ɪn ˈeɪʃə/ The magician showed me the way to the station. /ðə məˈdʒɪʃn̩ ʃəʊd miː ðə ˈweɪ tə ðə ˈsteɪʃn̩/ She sells seashells on the seashore. /ʃi selz ˈsiːʃelz ɒn ðə ˈsiːʃɔː/ Shy Shelly says she shall sew sheets. /ʃaɪ ˈʃeli ˈsez ʃi ʃə səʊ ʃiːts/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

202

Consonants

AI 4.14 The spellings of RP /ʒ/. Further practice genre /ˈʒɑːnrə/ leisure /ˈleʒə/ garage /ˈɡærɑːʒ/

vision /ˈvɪʒn̩/ massage /ˈmæsɑːʒ/ usual /ˈjuːʒʊəl/

That television is unusual. /ðət ˈtelɪˌvɪʒn̩ z ʌnˈjuːʒʊəl/ My pullover is beige and orange. /maɪ ˈpʊləʊvə z beɪʒ ənd ˈɒrɪndʒ/ You must take some decisions. /ju məst teɪk səm dɪˈsɪʒn̩z/ 4.2.2.5 RP glottal fricative /h/ IPA h x ɦ

Symbols Lower-case H. Lower-case X. Hooktop H.

Identification /h/ Fortis voiceless glottal fricative. Allophones [ɦ] Voiced glottal fricative. Description To articulate RP /h/ the soft palate is in its raised position and there is friction at the glottis and through the oral cavity, but as the vocal folds do not vibrate there is no distinctive opposition with a voiced counterpart as in the other English fricatives. The glottal fricative is always found before vowel sounds and it may be regarded as a strong, voiceless onset to the subsequent vowel because it has the same articulation (i.e. position of the tongue, lips, soft palate and configuration of the pharynx) as that of the vowel following it. As a result, /h/ will exhibit different types of friction (patterns of resonance) depending on the quality of the subsequent vowel (e.g. he /hiː/, who /huː/, hard /hɑːd/). See the animation and video under RP /h/ (compared with Spanish /x/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations /h/ occurs only in word-initial and word-medial positions. It remains voiceless word-initially, but between voiced sounds it may have some voicing, the result

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

203

being a breathy vowel or a voiced glottal fricative [ɦ], especially when initial in accented syllables, as in ahead [əˈɦed], and behind [bɪˈɦaɪnd], and less frequently in unaccented syllables, as in anyhow [ˈeniɦaʊ] and boyhood [ˈbɔɪɦʊd]. Spellings Table 27: The spellings of RP /h/ /h/

help /help/, hard /hɑːd/ silent + (only before who /huː/, whole /həʊl/

Note, however, that the spelling may also be silent word-initially, as in such words as heir /eə/, hour /ˈaʊə/, honour /ˈɒnə/, honest /ˈɒnɪst/, which must therefore be preceded by the forms of the indefinite and definite articles used before vowels (/ən/, /ði/), but also word-medially, as in exhibit /ɪɡˈzɪbɪt/, exhilarate /ɪɡˈzɪləreɪt/, shepherd /ˈʃepəd/, vehement /ˈviːəmənt/, vehicle /ˈviːɪkl ̩/ (see § 7.4.1.2). Regional and social variants Initial /h/ tends to be dropped in basilectal81 forms of most accents of Britain and Australia, and as a result such minimal pair distinctions are lost as hill – ill, high – eye, hair – air, which are pronounced as [ɪl], [aɪ], and [eə], respectively (Hughes et al. 2013). Although such h-dropping tends to be socially unacceptable, in General RP function words frequently lose initial /h/ in unaccented, noninitial positions in connected speech (see § 5.8). It may also occur that a glottal stop replaces initial /h/ (see § 5.2.8.2), and so a hill may be pronounced as [ə ˈʔɪl] (Cruttenden 2014: 208). Comparison with Spanish and advice As Spanish does not have the glottal fricative as a distinctive phoneme, in order to avoid the usual mistake to pronounce a velar fricative [x] instead of [h] (García Lecumberri 1999, 2000), SSLE should make the effort of completely devoicing the following vowel, as in hen [e̥en]. Only in the series /hju/, as in huge /hjuːdʒ/, can a fricative noise also be made resulting from a stricture in the mouth (Finch and Ortiz Lira 1982: 70). 81 The term basilectal is applied to a linguistic variety that is most remote from the prestige language (the matrilect or acrolect) and may also be contrasted with the intermediate varieties (or mesolects) (Crystal 2008: 51).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

204

Consonants

Similarly, as already noted, a correct distinction should be made between those minimal pairs differing only in pronunciations with and without initial /h/ (hill – ill), while being well aware of the RP tendency to elide initial /h/ in unaccented function words in connected speech, as will be further discussed in Section 5.8 (Cruttenden 2014: 208). Further practice AI 4.15 The spellings of RP /h/. Further practice help /help/ behind /bɪˈhaɪnd/ how /ˈhaʊ/

who /huː/ happy /ˈhæpi/ heaven /ˈhaʊ/ˈhevn̩//

hip /hɪp/ hunt /hʌnt/

Hallo, how are you? Hungry? /həˈləʊ | ˈhaʊ ə ju | ˈhʌŋɡri |/ Are you hoping to stay at a hotel? /ə ju ˈhəʊpɪŋ tə steɪ ət ə ˌhəʊˈtel |/ How high is that hill? /ˈhaʊ haɪ z ðət hɪl |/ 4.2.3 Affricates: RP /tʃ dʒ/ vs. PSp /tʃ/ The manner of articulation of affricates is quite similar to that of plosives, taking place in three steps, as follows. Firstly, the articulators come into contact, in such a way that the active articulator is moved towards the passive articulator. Secondly, a complete closure is produced so that the lung air cannot go out of the mouth, but is stopped at the point in the mouth where the articulators meet; the soft palate is in its raised position, which means that the lung air cannot be released through the nose either. The vocal folds may or may not vibrate during this compression stage. Thirdly, the articulators are separated, and so the lung air goes out of the mouth slowly, producing a slight friction, which is not so strong as that of fricative sounds (see also e.g. Gimson 1984: 172–177; Roach 2005: 54; Collins and Mees 2003: 163–166; Collins and Mees 2009: 44–45, 80– 82; Cruttenden 2014: 186–191). Any plosive whose release stage is produced in this way and is articulated approximately at the point where the plosive is made may be called an affricate. In English, in addition to the weakening of plosives by affrication (discussed in § 5.2.8.1), eight affricate series result from this type of release: [tʃ dʒ tɹ dɹ ts dz tθ dð]. Of these only /tʃ dʒ/ are considered to have phonemic status because

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

205

they can be contrasted as sound complexes in all word positions with a change of meaning. The former posits an additional medial distinction between closeknit [tʃ ] (butcher /ˈbʊtʃə/) and disjunct [t] + [ ʃ ] (lightship /ˈlaɪtʃɪp/) realisations, friction being shorter in the former. Besides, both /tʃ dʒ/ have commutability restrictions. In the case of /tʃ/, the plosive element may be commutated within the same syllable with zero or /l/ (as in watch, wash, Welsh), and the fricative element, with zero and /s r j w/ (as in catch, cat, cats, and chip, tip, trip, tune, twin). For /dʒ/ the commutability is restricted to zero in word-medial positions, in the case of the plosive element (ledger, leisure), and to zero and /r j w z/, in the fricative (as in jest, dam, dressed, dune, dwell, and hedge, head, heads). With more commutability possibilities but a more restricted distribution excluding final positions, [tɹ] and [dɹ] are normally regarded as series of two separate sounds /t d/ + /r/, in which the plosives have retracted realisations [t̠] [d̠] and /r/ will have different realisations depending on its occurrence in close-knit or disjunct contexts. In close-knit sequences, involving two sounds within one syllable, /r/ is voiced and fricativised (or produced with audible friction) [dɹ̝], if preceded by /d/ (draft [dɹ̝ɑˑft], but it is devoiced and fricativised [ɹ̥] after /t/ (mattress [ˈmætɹ̥ɪs]); whereas in disjunct sequences, where there is a syllable boundary between the two sounds, /r/ is approximant and voiced ( footrest [ˈfʊtɹest]). /ts dz/ (cats [kæts], bids [bɪdz], outset [ˈaʊtset]) are more restricted as they do not occur in initial positions (unless in foreign words) and only doubtfully in close-knit medial contexts (curtsey [ˈkɜːtsi]); whereas /tθ dð/ are even more restricted, with very few word-final occurrences (eighth [eɪt ̪θ] width [wɪt ̪θ]). Acoustically, as already detailed in Section 2.4.3, affricates show the characteristics of their components: a break for the stop combined with a highfrequency noise for the fricative. IPA tʃ dʒ ʝ

Symbols T-Esh ligature. D-Ezh ligature. Curly-tail J.

Identification /tʃ/ Fortis voiceless palato-alveolar affricate. /dʒ/ Lenis voiced palato-alveolar affricate. /ʝ / Lenis voiced palatal fricative.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

206

Consonants

Allophones /tʃ/ [tʃ ]̬ ‘voiced’ [tʃʷ] ‘labialised’ [tʃʲ] ‘palatalised’

/dʒ/ [d̥ʒ]̊ ‘completely devoiced’ in word-final positions [d̥ʒ] ‘partially devoiced’ in word-initial positions [dʒʷ] ‘labialised’ [dʒʲ] ‘palatalised’

Description To articulate RP /tʃ/ the tip and blade of the tongue are in contact with the alveolar ridge and the front of the tongue with the hard palate. The lung air is stopped behind the complete closure formed by the two articulators and then it is released slowly, so friction is produced. As a fortis consonant, it is produced with more energy than its voiced counterpart and with the vocal folds wide apart, thereby reducing the length of previous vocalic and sonorant sounds as in porch [pɔˑtʃ ] and nature [ˈneˑɪtʃə]. /dʒ/ is pronounced in the same way as /tʃ/ but the vocal folds may vibrate during all or part of both the stop and fricative stages. As a lenis consonant, it is articulated with less muscular energy than its fortis counterpart. Aspiration does not take place in the production of this sound, nor does it shorten the length of preceding vocalic or sonorant sounds (large [lɑːdʒ]). See the animations and videos under RP /tʃ/ and /dʒ/ (compared with Spanish /tʃ/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations RP affricates can appear in all possible contexts, but they differ from plosives in that they never lose their (fricative) release stage. /tʃ/ remains voiceless initially (chin [tʃɪn]) and finally (catch [kætʃ ], but between voiced sounds it gains some voicing ( furniture [ˈfɜːnɪtʃə]). /dʒ/, on the ̬ other hand, is fully voiced between voiced sounds (pigeon [ˈpɪdʒən]), but it loses some of its voicing when it occurs word-initially (gender [ˈd̥ʒendə]), and it may be completely devoiced in word-final positions (village [ˈvɪlɪd̥ʒ]̊ ). Another set of allophonic realisations involve the position of the lips, with lip-rounding (choose [tʃ ʷuːz], job [dʒʷɒb]) or lip-spreading (cheese [tʲiːz], village [ˈvɪlɪdʒʲ]) in the vicinity of rounded and spread vowels or semi-vowels, respectively (see § 5.2.3). Lastly, in such consonant clusters as /ntʃ/ and /ndʒ/ occurring in word-final positions the medial stop may be omitted (lunch [lʌnʃ ], strange [streɪnʒ]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

207

Spellings Table 28: The spellings of RP /tʃ dʒ/ /tʃ/

/dʒ/

tich /tɪtʃ/ child /tʃaɪld/

/ general /ˈdʒenrəl/ college /ˈkɒlɪdʒ/

match /mætʃ/

/ bridges /ˈbrɪdʒɪz/ bridge /brɪdʒ/

future /ˈfjuːtʃə/

conjunction /kənˈdʒʌŋkʃn̩/

, question /ˈkwestʃən/ righteous /ˈraɪtʃəs/

adjacent /əˈdʒeɪsənt/

Rare cases (Loanwords)

cello /ˈtʃeləʊ/

avalanche /ˈævəlɑːntʃ/

Rare cases procedure /prəˈsi:dʒə/ soldier /ˈsəʊldʒə/ exaggerate /ɪɡˈzædʒəreɪt/ nostalgia /nɒˈstældʒə/

Regional and social variants In RP there exist fluctuations between [tʃ dʒ] and [tj] [dj] pronunciations in the onset of both unaccented (culture [ˈkʌltʃə] – [ˈkʌltjə], soldier [ˈsəʊldʒə] – [ˈsəʊldjə]) and accented syllables (tune [tʃuːn] [tjuːn] – dune [dʒuːn] [djuːn]). [tj] [dj] pronunciations are associated with careful, Conservative RP, whereas the [tʃ dʒ] pronunciations are well established in RP and they tend to be preferred by younger speakers. No such alternations occur in most accents of GA because they have lost the /j/ element following /t d/ sequences, which are realised as [tuː] [duː] (tune [thuːn] – dune [duːn]), and so neither [tj] [dj] nor [tʃ dʒ] pronunciations are heard. In addition, some RP speakers omit the stop element in the clusters [ntʃ ] [ndʒ] in word-final positions (French [frenʃ ], strange [streɪnʒ̊]), while [dʒ] is increasingly simplified to [ʒ] (liege [liːʒ̊]), in accented-syllable final positions, and occasionally to [d], in unaccented positions (dangerous [ˈdeɪndrəs]) (Finch and Ortiz Lira 1982: 71; Cruttenden 2014: 190–191). Comparison with Spanish and advice As /tʃ/ has phonemic status in Spanish, but its voiced counterpart does not, the main difficulty for SSLE involves /dʒ/. In Latin-American Spanish (Argentina,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

208

Consonants

Uruguay and Chilean) the voiced palatal fricative /ʝ/ is realized as a fricative [ʒ] or an affricate [dʒ] as a result of “žeísmo”, especially in spellings after /n/ (cónyuge ‘partner’) and /l/ (el yate ‘the yacht’), as well as in emphatic pronunciations of some words (yo ‘I’ [dʒo], oye ‘listen’ [odʒe]). [dʒ] is also used in Catalan in such words as injust [ɪnˈdʒust], although in English the sound has a much freer distribution. Besides, among younger Spanish speakers the phenomenon of “šeísmo” has been detected, whereby [ ʝ] is devoiced to [ ʃ ] or, in emphatic pronunciations, it is devoiced and affricated to [ʧ]. Likewise, in Andalusian accents /tʃ/ may be replaced by [ ʃ ]. It should be noted, however, that “lleísmo” [ʎeˈizmo], or the pronunciation that distinguishes between /ʝ/ (spelled “y”) and the palatal lateral approximant /ʎ/ (spelled “ll”) is attested sparingly in the dialects of Spanish (broadly, restricted to parts of Northern Spain and of the Andean highlands of Ecuador, Paraguay and Bolivia) and is a recessive feature of modern Spanish. Most varieties of Spanish (most of Spain, Argentina, Chile, Venezuela, Perú, Ecuador and ͡ Colombia) are characterized by “yeísmo” [ɟʝeˈizmo]. This is a pronunciation in which /ʝ/ and /ʎ/ are merged into the former phoneme with the resulting loss of the palatal lateral approximant, so that and spellings represent the same sound /ʝ/, usually realized as a palatal approximant (in word initial, as in yeso ‘plaster’ [ˈʝeso], and intervocalic position, as in hoyo ‘hole’ [ˈoʝo]) or affricate [ɟ] (after a nasal or lateral consonant, as in un yeso ‘a plaster’ [uń ˈɟeso], or in emphatic pronunciations, as in yo ‘I’ [ɟo]) (Avezado 2009). From the aforementioned it can be concluded that, although for some SSLE the pronunciation of [dʒ] should pose little or no difficulty because this sound exists in their phonetic inventories, for most SSLE, who do not have it in their native Spanish accents, it may be hard to reproduce. As a consequence, the English voiced palatal affricate ( jet /dʒet/) may be mispronounced as a as a palatal lateral approximant (*[ˈʎet]), as a voiced palatal fricative (*[ˈʝet]), or affricate (*[ˈɟet]), or very frequently by a palatal approximant (*[jet], enjoy *[ɪnjɔɪ], just *[jʌst]), based mainly on the wrong assumption that spellings have to be pronounced as [j]. These mispronunciations create communication problems that result from the neutralisation of the /dʒ/ – /j/ contrast that exists in English between such words as jet /dʒet/ – yet /jet/ and jot /dʒɒt/ – yacht /jɒt/. When affricates come together in connected speech, the plosive element of the first affricate may be omitted (Dutch cheese [dʌʃ ˈtʃiːz], change gestures [ˈtʃeɪnʒ ˈdʒestʃəz], lounge chair [ˈlaʊnʒ ˈtʃeə]). In these cases, especial care should be taken not to omit the fricative element of the first affricate as pronunciations with such omissions would be judged as unacceptable. Special attention should also be paid to the /tr/ and /dr/ sequences because of the retracted place of articulation of the plosive and the fricative or devoiced

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

209

(when following /t/) quality of /r/, taking care not to confuse them with /tʃ/ (trees – cheese [t̠hɹ̥jiːz] – [tʃ jiːz]) and /dʒ/ (draw – jaw [d̠̥ɹwɔː] – [d̥ʒwɔː]), respectively. Similarly, in order to distinguish [tʃ ] and [d̥ʒ]̊ in word-final positions, SSLE should remember to shorten the length of the vowel preceding [tʃ ] (search – surge [sɜˑtʃ ] – [sɜːd̥ʒ]̊ ). Production and discrimination exercises are required to distinguish in all possible contexts /tʃ/ and /dʒ/, as well as /tʃ dʒ/ and [ ʃ ʒ]. A good starting point could be to pronounce the [tj] [dj] sequences but moving the tongue forward (Finch and Ortiz Lira 1982: 69; Pennock Speck 2001; Gallardo del Puerto 2005: 202; Estebas Vilaplana 2009: 67–70; Cruttenden 2014: 191–192). Further practice AI 4.16 The spellings of RP /ʧ/. Further practice chin /tʃɪn/ lecture /ˈlektʃə/ watch /wɒtʃ/

choose /tʃuːz/ kitchen /ˈkɪtʃɪn/ teach /tiːtʃ/

cheese /tʃiːz/ future /ˈfjuːtʃə/

He teaches Dutch. /hi ˈtiːtʃɪz dʌtʃ/ I like cheese, cherries and chips very much. /ˈaɪ ˈlaɪk tʃiːz | ˈtʃerɪz ənd tʃɪps ˈveri ˈmʌtʃ/ There’s a chicken in the kitchen. /ðeəz ə ˈtʃɪkɪn ɪn ðə ˈkɪtʃɪn/ If Stu chews shoes, should Stu choose the shoes he chews? /ɪf ˈstuː tʃuːz ʃuːz | ʃəd ˈstuː tʃuːz ðə ʃuːz hi tʃuːz/

AI 4.17 The spellings of RP /ʤ/. Further practice gin /dʒɪn/ ledger /ˈledʒə/ badge /bæʤ/

generous /ˈdʒenərəs/ adjacent /əˈdʒeɪsənt/ college /ˈkɒlɪdʒ/

conjunction /kənˈdʒʌŋkʃn̩/ bridge /brɪdʒ/

Germans are generally generous. /ˈdʒɜːmənz ə ˈdʒenrəli ˈdʒenərəs/ Just sit on the edge of the bridge. /dʒʌst sɪt ɒn ði edʒ əv ðə brɪdʒ/ A large majority voted for the judge. /ə lɑːdʒ məˈdʒɒrɪti ˈvəʊtɪd fə ðə dʒʌdʒ/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

210

Consonants

4.2.4 Nasals: RP /m n ŋ/ vs. PSp /m n ɲ/ In RP there are three nasal phonemes /m n ŋ/. Articulatorily, the manner of articulation of nasals (also known as nasal stops) resemble that of plosives (or oral stops) in that in both groups an active articulator comes into contact with a passive articulator so that a complete closure of the oral chamber is made. The main difference between oral and nasal stops is the position of the soft palate. While in plosives it is raised, and so the airstream escapes through the mouth, in nasals it is lowered so that the lung air is unimpededly released through the nasal cavity, for which they are regarded as continuants, giving the sound the resonance provided by the nasopharyngeal cavity. Nasals are also referred to as frictionless continuants, because no audible friction is produced in their articulation, as it occurs with other continuants such as fricatives. In addition, nasals are considered to be vowel-type sounds because they are normally voiced, which means that the vocal folds do vibrate in the compression stage. As a consequence, they neither present remarkable voiced/voiceless oppositions, nor reduce the length of the previous sound. Another feature that evidences the vocalic nature of nasals is their capability to become syllabic, frequently /n/ (eleven [ɪˈlevn̩]), less frequently /m/ (happen [ˈhæpm̩ ]) and occasionally /ŋ/ (thicken [ˈθɪkŋ̍]), thereby conveying the syllabic prominence that characterises vowels, as already explained in Section 1.3.3.3. As regards place of articulation, English nasals and plosives have again the same areas of articulation. /m/ is bilabial (like /p b/), the two articulators being the lower lip (active articulator) and the upper lip (passive articulator); whereas /n/ is alveolar (like /t d/) and /ŋ/ velar (like /k g/), the two articulators being in the former case the tip of the tongue (active articulator) and the alveolar ridge (passive articulator), and in the latter case the back of the tongue (active articulator) and the soft palate (passive articulator). As a consequence, if the nasal passage is blocked, as when having a cold, nasals (manner /ˈmænə/) may be realised as voiced plosives ([ˈbædə]). /n/ shows the highest number of allophonic realisations because, as already explained about other alveolar consonants, its place of articulation is particularly liable to be affected by that of the following consonant. Turning to their main acoustic features, as also detailed in Section 2.4.3, nasals are characterised by the absence of energy around 1,000 Hz, the presence of a low-frequency resonance or “murmur” below 500Hz, and abrupt transitions from and into the neighbouring sounds (see also e.g. Gimson 1984: 193–198; Roach 2005: 58–61; Collins and Mees 2003: 167–168; Collins and Mees 2009: 54–57, 84–87; Cruttenden 2014: 209–217).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

IPA m n ŋ

211

Symbols Lower-case M. Lower-case N. Eng.

Identification /m/ lenis voiced bilabial nasal. /n/ lenis voiced alveolar nasal. /ŋ/ lenis voiced velar nasal. Allophones [m] [m̥ ] ‘devoiced’ [mʷ] ‘labialised’ [mʲ] ‘palatalised’ [m̩ ] ‘syllabic’ [ɱ] ‘labio-dental’

[n] [n̥ ] ‘devoiced’ [nʷ] ‘labialised’ [nʲ] ‘palatalised’ [n̩] ‘syllabic’ [ɱ] ‘labio-dental’ [ṉ] ‘retracted’ ‘post-alveolar’ [n̟] ‘advanced’ [n̪] ‘dental’ [m] ‘bilabial’ [ŋ] ‘velar’

[ŋ] [ŋ̊] ‘devoiced’ [ŋʷ] ‘labialised’ [ŋʲ] ‘palatalised’ [ŋ̍] ‘syllabic’ [ŋ̄ ] ‘retracted’ [ŋ] ‘advanced’

Description In order to pronounce /m/ the two lips are in contact forming a closure, whereas to produce /n/ and /ŋ/ the closure is made between the tip of the tongue and the alveolar ridge, and the back of the tongue and the soft palate, respectively. Simultaneously, in all three cases the soft palate is in its lowered position so that the lung air goes out through the nose, adding the resonance of the nasal cavity to those of the oral and pharyngeal cavities. As voiced phonemes, nasals are produced with varying degrees of vocal fold vibration and they do not shorten the length of preceding vocalic or sonorant sounds (bean [biːn]). See the animations and videos under RP /m/, /n/ and /ŋ/ (compared with Spanish /m n/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations While /m n/ can occur in all positions, /ŋ/ is restricted to syllable codas, after short vowels /ɪ æ ɒ ʌ e/, either word-medially or word-finally. Phonologically, Roach (2005: 59–60) explains that it is difficult to work out a rule as to how to

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

212

Consonants

pronounce word-medial and word-final spellings: either as /ŋ/ or as /ŋk/ or /ŋɡ/ combinations, respectively. Roach suggests that spellings should always be pronounced as /ŋk/ (linking /ˈlɪŋkɪŋ/, sink /sɪŋk/), whereas word-medial may have either /ŋ/ pronunciations, if the digraph occurs at the end of a morpheme (singing /ˈsɪŋɪŋ/, singer /ˈsɪŋə/, hanger /ˈhæŋə/), or /ŋɡ/ pronunciations, if it appears in the middle of a morpheme (anger /ˈæŋɡə/, finger /ˈfɪŋɡə/). When the spelling occurs in word-final position, [ɡ] is never pronounced (long /lɔŋ]). There exist, however, two important exceptions to these tendencies. Comparative and superlative forms are pronounced with medial /ŋg/, instead of /ŋ/ (longer /ˈlɒŋɡə/, longest /ˈlɒŋɡɪst/), despite consisting of two morphemes and having morpheme-final , and therefore they “behave” as single-morpheme words. We have seen that these three nasals are generally voiced. But, as occurs with other voiced phonemes, they are only fully voiced between voiced sounds (amber [ˈæmbə], annoy [əˈnɔɪ], finger [ˈfɪŋɡə]). Devoiced allophones may be heard in word-initial and word-final positions (mummy [ˈm̥ ʌmi], autumn [ˈɔːtəm̥ ], bacon [ˈbeɪkən̥], hang [hæŋ̊]), as well as when nasals follow a voiceless consonant (smart [sm̥ ɑːt], sneeze [sn̥ iːz], bacon [ˈbeɪkŋ̊ˈ]).])) In addition to their syllabic realisations, which, as already noted, can occur in the three nasals (rhythm [ˈrɪðm̩ ], cotton [ˈkɒtn̩], taken [ˈteɪkŋ̍]), although /n/ records the highest incidence, both /m/ and /n/ may have labio-dental realisations [ɱ], if they are followed by the labio-dental fricatives /f/ or /v/ either within words or in close-knit word-boundaries (comfort [ˈkʌɱfət], circumvent [ˌsɜːkəɱˈvent], infant [ˈɪɱfənt], envy [ˈeɱvi]). Also word-medially or at wordjunctures /n/ may have a dental realisation when it occurs before and after dental fricatives (ninth [naɪn̪θ]) and a post-alveolar allophonic when followed by /r/ (inroad [ˈɪṉɹəʊd]). Likewise, [ŋ] and [m] can be two other allophonic variants of /n/, when it is followed by either a velar (ten coins [teŋ ˈkɔɪnz]) or a bilabial consonant (in bed [ɪm ˈbed] in close-knit word boundaries, respectively; and [m] may also be heard when word-final /n/ is preceded by bilabial /p b/ (happen [ˈhæpm̥̩ ], carbon [ˈkɑːbm̥̩ ]. The last set of allophonic realisations that affect the three nasals involve the position of the lips, with lip-rounding (mob [mʷɒb], no [nʷəʊ], long [lɒŋʷ]) or lip-spreading (mule [mʲjuːl], mining [ˈmaɪnʲɪŋʲ]) in the vicinity of rounded and spread vowels or semi-vowels, respectively (see § 5.2.3).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

213

Spellings Table 29: The spellings of RP /m n ŋ/ /m/

/n/

/ŋ/

mother /ˈmʌðə/ damage /ˈdæmɪdʒ/

plan /plæn/ bend /bend/

sing /sɪŋ/ longing /ˈlɒŋɪŋ/

mummy /ˈmʌmi/ immense /ɪˈmens/

planning /ˈplænɪŋ/ connect /kəˈnekt/

(+ /k/ or /g/) sink /sɪŋk/ bangle /ˈbæŋɡl ̩/

+ silent bomb /bɒm/, climb /klaɪm/

silent + benign /bəˈnaɪn/, gnaw /nɔː/ silent + know /nəʊ/ acknowledge /əkˈnɒlɪdʒ/

Rare cases + silent solemn /ˈsɒləm/, column /ˈkɒləm/ silent + calm /kɑːm/, balm /bɑːm/ silent + diaphragm /ˈdaɪəfræm/

Rare cases silent + mnemonic /nɪˈmɒnɪk/ silent

+ pneumatic /nju:ˈmætɪk/ pneumonia /njuːˈməʊnɪə/

Rare cases

tongue /tʌŋ/

Regional and social variants Neither /m/ nor /n/ have important social or regional variants. The most remarkable feature concerns the alternations between [ŋ] and [ŋg] pronunciations where RP would have /ŋ/ (e.g. singing [ˈsɪŋɪŋ] – [ˈsɪŋgɪŋg], which are typical of the north-west of England (e.g. Lancashire, Derbyshire, Cheshire and Staffordshire). Comparison with Spanish and advice English /n/ and /ŋ/ pose a problem for speakers of Spanish as regards their phonemic status. In Spanish [ŋ] occurs as an allophone of [n] before velars (e.g. hongo ‘fungus’ [ˈoŋɡo]). Its occurrence without an intervening velar is only registered in Central American Spanish, as well as in South American Spanish (Peruvian, Ecuadorian), where [ŋ] is used as a common realisation of /n/. The most common error for SSLE is to produce [n] instead of [ŋ], and so English gerunds, for example, such as linking [ˈlɪŋkɪŋ], watching [ˈwɒtʃɪŋ], playing [ˈpleɪɪŋ] and coming [ˈkʌmɪŋ], may be pronounced with a final [n], as in *[ˈlɪŋkɪn], *[wɒtʃɪn], *[pleɪɪn] and *[ˈkʌmɪn], which may cause intelligibility problems derived from the loss of the [ŋ] – [n] phonemic contrast that exists in English (ran /ræn/ – rang /ræŋ/).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

214

Consonants

Turning to /m/, whereas in English it may occur in all possible contexts, in Spanish its occurrence in word-final positions is highly restricted, since cases like álbum are exceptional and the final nasal is often pronounced as an alveolar *[alβun] rather than a bilabial. SSLE should therefore be aware of the need to pronounce word-final /m/ appropriately, keeping in mind the distinct phonemic nature of English /n/ and /ŋ/: the former may occur in all possible contexts, and the latter word-medially or word-finally, whether or not it is next to another velar (Finch and Ortiz Lira 1982: 64, 69; Gallardo del Puerto 2005; Estebas Vilaplana 2009: 82–85). Further practice AI 4.18 The spellings of RP /m/. Further practice met /met/ simmer /ˈsɪmə/ Pam /pæm/

month /mʌnθ/ amber /ˈæmbə/ album /ˈælbəm/

bomb /bɒm/ someone /ˈsʌmwʌn/

Mummy’s photo was in the album all the time. /ˈməmiz ˈfəʊtəʊ wəz ɪn ði ˈælbəm ɔːl ðə ˈtaɪm/ We’ll discuss it in the forum. /wil dɪˈskʌs ɪt ɪn ðə ˈfɔːrəm/ Now it’s the month of May, it’s getting warm. /naʊ ɪts ðə mʌnθ əv meɪ ɪts ˈɡetɪŋ wɔːm/

AI 4.19 The spellings of RP /n/. Further practice net /net/ sinner /ˈsɪnə/ pen /pen/

never /ˈnevə/ find /faɪnd/ none /nʌn/

know /nəʊ/ ant /ænt/ plan /plæn/

No, it’s never nice there. /nəʊ | ɪts ˈnevə naɪs ðeə/ No news is good news. /nəʊ njuːz ɪz ɡʊd njuːz/ They found a nice new nanny for the baby. /ˈðeɪ faʊnd ə naɪs njuː ˈnæni fə ðə ˈbeɪbi/ Nine nice night nurses nursing nicely. /naɪn naɪs naɪt ˈnɜːsɪz ˈnɜːsɪŋ ˈnaɪsli/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

215

AI 4.20 The spellings of RP /ŋ/. Further practice singer /ˈsɪŋə/ ping /pɪŋ/ sink /sɪŋk/

tongue /tʌŋ/ uncle /ˈʌŋkl ̩/ hungry /ˈhʌŋɡri/

angry /ˈæŋɡri/ thinking /ˈθɪŋkɪŋ/

I heard a loud bang, and someone singing a song. /ˈaɪ hɜːd ə laʊd bæŋ | ənd ˈsʌmwʌn ˈsɪŋɪŋ ə sɒŋ/ I’m going to the bank. /aɪm ˈɡəʊɪŋ tə ðə bæŋk/ The singer, when he was younger, broke his finger. /ðə ˈsɪŋə | wen hi wəz ˈjʌŋɡə | brəʊk ɪz ˈfɪŋɡə/ 4.2.5 Approximants This section focuses on four RP consonantal phonemes /l r j w/ that share the same manner of articulation described as open approximation (see also Gimson 1984: 193–198; Roach 2005: 58–61; Collins and Mees 2003: 168–171; Collins and Mees 2009: 54–57, 84–87; Cruttenden 2014: 209–217). This means that these four sounds require a constriction which is typically greater than that required for a vowel, but they are produced with an approximation of the articulators which is too open to cause any friction (unlike that of fricatives, which is characterised as close approximation), for which they (together with vowels)82 are also referred to as frictionless continuants (see 2.3.2.2). /r j w/ are characterised as central approximants in that the airstream passes straight along a central/median/mid-sagittal channel through the vocal tract, whereas /l/ is the only lateral approximant of the RP phoneme inventory, because the air flows across the lowered side rims of the tongue around some central obstruction, so that the action of the side rims against the upper molars can be felt, as well as the contact made between the tongue tip and the alveolar ridge (Ashby 2011: 50). Accordingly, in what follows it will be shown that, considering manner of articulation, /l/ is categorised as alveolar, whereas /r/, /j/ and /w/ will be classified as post-alveolar, palatal and labial-velar, respectively. Distributionally, approximants behave very similarly. They are the second and third elements of two of the most recurrent initial consonant clusters in English: C + /l r w j/ and /s/ + /p t k/ + /l r j w/. In both contexts approximants 82 In some analyses, [h] is also considered an approximant as the voiceless equivalent of the vowel following (Crystal 2008:32).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

216

Consonants

are devoiced, if they are preceded by a voiceless consonant (clay [kl ̥eɪ], try [tɹ̥aɪ], quick [kw̥ ɪk], fuel [ˈfȷ ̊uːəl], spray [spɹ̥ei], stewed [stȷ ̊uːd]), and they are shortened, like long vowels and vocalic glides, if followed by a voiceless consonant (so compare kilt [kɪɫˑt] with killed [kɪɫːd]). Acoustically, as already detailed in Section 2.4.3, approximants exhibit a voicing bar along with formant-like structures that resembles those of vowels, and show the following most remarkable characteristics (O’Connor et al. 1957; Cruttenden 2014: 225, 220): (1) /r/ may be identified in spectrograms by its steepingly rising transitions (for F1, F2, F3 and F4) to a following vowel. (2) For most realisations of /l/ F3 is high and weak (around 2500 Hz) and transitions from and to vowels tend to be slower than those for nasals although faster than those for glide consonants. (3) /j/ and /w/ have a two- or three-formant structure similar to that of [iː] and [uː] respectively, and their steady state is even shorter than that of the other two approximants (around 30 msecs). In order to be able to devoice /l r j w/, SSLE could try to produce them as if they were whispered. It should be kept in mind that, as a side-effect of aspiration, this devoicing effect is blocked in two contexts: (1) when an /s/ precedes the consonant cluster; and (2) when the syllable is not stressed. So compare, cream [khɹ̥iːm] and plastic [ˈpl ̥æstɪk], which show fortis plosive aspiration and approximant devoicing, with scream [skɹiːm] and plasticity [plæˈstɪsɪti], where there is neither aspiration nor approximant devoicing (Estebas Vilaplana 2009: 100–101).

4.2.5.1 Liquids: RP /r l/ vs. PSp /r ɾ l ʎ/ /l/ and /r/ are grouped together because both are liquid approximants, that is, both are apico-alveolar approximant sounds that are made by placing the tip of the tongue on the alveolar ridge (cf fn 5) (see also Gimson 1984: 193–198; Roach 2005: 58–61; Collins and Mees 2003: 168–172; 176–181; Collins and Mees 2009: 54–57, 84–87; Cruttenden 2014: 209–217; Walsh-Dickey 1997). IPA r ɹ ɾ l

Symbols Lower-case R. Turned R. Fish-hook R. Lower-case L.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

217

Identification /r/ [ɹ] Voiced post-alveolar approximant. /r/ Voiced post-alveolar trill. /ɾ/ Voiced post-alveolar tap. /l/ Voiced alveolar lateral approximant. Allophones /r/ [ɹ] ‘post-alveolar approximant’ [ɹˠ] ‘velarised alveolar approximant’ [ɻ ] ‘retroflex’ [ɾ] ‘alveolar tap’ [ɹ]̝ ‘voiced fricative’ [ɹ]̥ ‘devoiced’ / ‘voiceless fricative’ [ɹʷ] ‘labialised’ [ɹʲ] ‘palatalised’ [ʋ] ‘voiced labio-dental approximant’ [ʀ] ‘voiced uvular trill’ [ʁ] ‘voiced uvular fricative fricative’

/l/ [l] ‘clear’ or ‘palatal’ [ɫ] ‘dark’ or ‘velar’

[l ̥] [ɫ ̥] ‘devoiced’ [lʷ] ‘labialised’ [lʲ] ‘palatalised’ [ɫ ]̪ ‘dental’ [ɫ ]̠ ‘retracted’ or ‘post-alveolar’ [ɫ ]̃ [l]̃ ‘nasalised’

Description For the articulation of RP /r/, the nasal resonator is shut off by raising the soft palate. The tip of the tongue is slightly retroflected close to, but does not touch, the rear part of the alveolar ridge, while the centre of the tongue is hollowed and the back rims are touching the upper molars, no contact being made with the roof of the mouth at any time. The vocal folds do vibrate and no friction is produced when the lung air goes out of the mouth. Although lip position is largely determined by that of the following vowel, English speakers tend to pronounce this sound with rounded lips (Roach 2005: 63). /l/, on the other hand, is pronounced with a complete median closure of the oral chamber by placing the tip (and blade) of the tongue in the alveolar ridge, and a partial lateral closure, that is, an opening on one or both sides of the mouth which is not radical enough to produce friction. As the soft palate is in its raised position shutting off the nasal resonator, the lung air is released laterally on both sides, or on one side, if there exists a unilateral closure between the side-rims and the upper side teeth. As it is also produced with vibration of the vocal folds, /l/ is classified as a voiced phoneme and therefore it does not affect the length of the previous sound ( feel [fiːɫ]). In addition, to pronounce dark [ɫ] the back of the tongue is raised in the direction of the soft palate, producing a velarised resonance close to that of back vowels, ranging in closeness from [ʊ], [ö] or [ɤ] to [ɔ] or [ʌ]; whereas the

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

218

Consonants

clear [l] variant is characterised by a front vowel resonance, which results from raising the front of the tongue in the direction of the hard palate. These articulatory differences have acoustic correlates in that the values of F1 (low) and F2 (weak) are quite far apart (around 300 and 1600 Hz) for clear [l], whereas they are close together (around 400 and 600 Hz) for dark [ɫ]. See the animations and videos under RP /r/ and /l/ (compared with Spanish /r ɾ l/) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations Since RP is a non-rhotic accent /r/ is normally pronounced only syllableinitially before vowels or glides (red /red/, arrow /ˈærəʊ/) (see fn 8). As a result, it does not sound when it is followed by a consonant or when it occurs in syllable-final or word-final positions ( far /fɑː/, farm /fɑːm/). However, when a word ends in the spelling and the following word begins with a vocalic sound, then /r/ is pronounced as an instance of so-called linking /r/ (He wore a suit /hi ˈwɔːr ə ˈsuːt /, far away /fɑːr əˈweɪ/, there are /ðər ˈɑː/) (see § 5.5). Sometimes speakers use a linking /r/ in words where there is no spelling (He saw a car /hi ˈsɔːr ə kɑː/), in which case we speak of intrusive /r/. However, intrusive /r/ should be avoided as it indicates an uneducated pronunciation. A retroflexed [ɻ ], curling the tip of the tongue backwards towards just behind the alveolar ridge, is often heard both before and between vowels (road [ɻəʊd], far off [fɑːɻ ɒf]), very [ˈveɻi])), in which case the vowel is said to be r-coloured or rhotacised (see fn 64) and the degree of retroflexion varies depending on regional, social and stylistic factors (Wells 1982: 432). /r/ may also be realised as a voiced alveolar tap or flap [ɾ], and so a short single tap being made with the tip of the tongue on the alveolar ridge, in intervocalic positions (very [ˈveɾi]), or the dental ridge, before /θ/ and /ð/ (three [θɾiː], thread [θɾed], with rage [wɪð ɾeɪdʒ]), the central hollowing of the tongue being retained. Both within words and between close-knit word-boundaries, /r/ has a fricative allophone, which is voiced when it follows /d/ [dɹ]̝ (draft [dɹ̝ɑˑft], bedroom [ˈbedɹ̝uːm], wide range [waɪd ɹ̝eɪndʒ]) and shows different degrees of devoicing when it follows a fortis plosive (upright [ˈʌpɹ̥aɪt], try [tɹ̥aɪ], cream [kɹ̥iːm]). Likewise, devoiced realisations will occur when /r/ follows a voiceless consonant ( friar [ˈfɹ̥aɪə], necessary [ˈnesəsɹ̥i], surf-riding [ˈsɜːfɹ̥aɪdɪŋ]), as well as in word-initial positions (red [ɹ̥ed]). Considering the position of the lips, variants with tight lip-rounding (room [ɹʷu:m]) or some degree of lip-spreading (read [ɹʲi:d]) will also occur in the vicinity of rounded and spread vowels or semi-vowels, respectively. Turning to /l/, it can occur in all positions in the word, but it has two allophonic variants depending on its place of occurrence. It is clear [l] before vowels or before palatal /j/ (love [lʌv], peculiar [pɪˈkjuːlɪə]), and it is dark [ɫ]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

219

before consonants and in final position (milk [mɪɫk], mill [mɪɫ]. Dark l [ɫ] has the effect of retracting and lowering the articulation of the preceding front vowel, so that [i:ɫ] pronunciations, for example, usually have a centring glide and for many RP speakers there is no contrast with [ɪəɫ] (reel – real) and they can pronounce words such as feel either as [fi:ɫ] or [fɪəɫ]. Similarly, before [ɫ] the second elements of vowel glides tends to be dropped or is otherwise reduced to [ə] or [ʊ] ( failed [feəɫd], aisle [aəɫ], owl [aʊ ɫ], boils [bɔəɫz], cruel [ˈkruːəɫ]), as already noted in Section 3.3. Both clear and dark l are voiced, but they are devoiced after voiceless consonants ( flee [fl ̥iː], bottle [ˈbɒtɫ ̥].̩ Moreover, when preceded by /p/ or /k/ in a stressed syllable (play [pl ̥eɪ], climb [kl ̥aɪm]), the devoicing of /l/ sounds like a kind of aspiration of the plosive. Adjacent to nasals, /l/ is nasalised [ɫ ]̃ (signalman ̃ [ˈsɪɡnɫmən]), and considering the place of articulation, as already seen in the description of other alveolar sounds, [ɫ] tends to approach that of the following consonant, both within words and in close-knit word-boundaries. Hence, if [ɫ] occurs next to /θ ð/ it will have a dental articulation [ɫ ]̪ (health, feel that [heɫ ̪θ fiːɫ ̪ ðæt]), and it will be post-alveolar [ɫ ]̠ before /r/ (already [ɔːɫ ̠ˈɹedi], all roses [ɔːɫ ̠ˈɹəʊzɪz]). Lip position is, as usual, another parameter of allophonic variation so that allophones with either lip-rounding (loo [lʷu:], cool [ku:ɫʷ]) or lip-spreading (lea [lʲi:], eel [i:ɫj]) will occur in the vicinity of rounded and spread vowels or semi-vowels, respectively. Under the conditions explained in Section 1.3.3.2, both /r/ and particularly /l/ (mostly dark) may be syllabic (generous [ˈdʒenɹ̩əs], memory [ˈmemɹ̩i], petal [ˈpetɫ̩], bottle [ˈbɒtɫ̩]). Spellings Table 30: The spellings of RP /l r/ /r/

/l/

round /ˈraʊnd/, staring /ˈsteərɪŋ/

(including ) let /let/, feel /fɪl/, middle /ˈmɪdl ̩/

starring /ˈstɑːrɪŋ/, sorry /ˈsɒri/

fill /fi:l/, allow /əˈlaʊ/

silent + write /ˈraɪt/, wreath /riːθ/

Rare cases silent + isle /aɪl/

+ silent rhyme /raɪm/ rhythm /ˈrɪðəm/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

220

Consonants

Note that there are words that are spelt with a single in British but normally have a double l in American English (dial(l)ing, distil(l), enrol(l), fulfil(l), signal(l), instal(l)ing, travel(l)ing) whereas in the following cases spellings are always silent. calm could folk

/kɑːm/ /kʊd/ /fəʊk/

half salmon should

/hɑːf/ /ˈsæmən/ /ʃʊd/

talk would

/ˈtɔːk/ /wʊd/

Regional and social variants Originating in basilectal London accents and then spreading to the Southeast and further away (e.g. Derby, Milton Keynes, Middlesborough), the labialisation of /r/ is so extreme that there exists a widespread tendency to substitute it (and also /w/) by /ʋ/, a voiced labio-dental approximant, although this is also a feature of early child language and has often been considered as a speech defect in adults (Cruttenden 2014: 225; Foulkes and Docherty 2000). In addition, either a voiced uvular trill [ʀ] or a voiced uvular fricative [ʁ] are residual r-variants of the north-east of England and Lowland Scottish English. Unlike in RP, in rhotic English accents (most American, Scottish, West England and Irish English and much of the rural areas of the South and Southwest of England), the phoneme /r/ is pronounced both in syllable initial and syllable-final positions, and so a distinction is made between such words as court-caught and poor-paw which are homophones in RP. Broadly speaking, in these rhotic accents prevocalic /r/ is generally pronounced as a velarised alveolar continuant [ɹ] ([ɹˠ]), but postvocalic retroflex [ɻ ] (bird [bɜːɻd], farm [faːɻm], lord [lɔːɻd]) is increasingly frequent and may become the mainstream pronunciation. Another feature of these accents resulting from their rhoticity is the absence of linking-/r/ and intrusive-/r/ (Wells 1982: 431–432; Hickey 2007: 320). Moving on to the lateral approximant, in the speech of London there exists a tendency either to omit dark [ɫ] or to substitute it by a vowel in the region of [ö], a centralised close-mid back rounded vowel, or [ɤ̈], its unrounded counterpart (sell [seö] – [seɤ̈]). Besides, the distributionally-based clear [l] and dark [ɫ] distintion that works in RP does not apply to other English accents. In SSE, the North of England and Wales, GA, Australian English and New Zealand English, dark [ɫ] is pronounced in all positions. In contrast, in South Wales and on Tyneside, as well as in West Indian English and southern Irish English clear [l] is found in all positions, although in the latter accent postvocalic dark [ɫ] may also be heard among young speakers (Wells 1982: 431; Hughes 2005: 114; Hickey 2007: 321; Cruttenden 2014: 220–221).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

221

Comparison with Spanish and advice Whereas RP is a non-rhotic accent in which /r/ is generally pronounced only before vowels and glides and is mostly realised as an approximant sound (no contact being made with the roof of the mouth), Spanish is a rhotic language which has two “r-phonemes” involving two different types of tongue-palate contact: tap [ɾ] (as in pero ‘but’), trill [r] (as in perro ‘dog’). These phonemes are contrastive only in intervocalic position. Their contrast is neutralised in postnuclear position (cardo ‘thistle’) (Quilis and Fernández 1996), elsewhere they occur in complementary distribution: [r] is used word-initially and after consonants, as well as an emphatic alternative to [ɾ] in word-final positions (Mott 2011: 258–259), although some experimental studies suggest a preference for a tap [ɾ] in all these contexts except for word-initially (Blecua 1999). In addition, both Spanish phonemes are described as retroflex, whereas in English not all instances of /r/ are retroflected. Hence a common mistake for SSLE is to use the two Spanish r sounds ([ɾ] or [r]) instead of English post-alveolar approximant /r/, which does not cause misunderstandings but gives a very strong foreign accent (García Lecumberri and Elorduy 1994; García Lecumberri 1999, 2000; Gallardo del Puerto 2005). To avoid such replacements and pronounce RP /r/ correctly, SSLE are advised to round their lips, but being careful not to exaggerate this gesture as too much rounding would make this consonant sound too much like [w], and try to produce a central vowel of the /ə/ /ɜː/ type. Likewise, to get as close as possible to native RP pronunciation, SSLE should not pronounce pre-consonantal and pre-pausal spellings, but rather should consider them as indications to lengthen or diphthongise the previous vowel(s) (Cruttenden 2014: 227–228). Nevertheless, in connected speech word-final spellings should normally be pronounced as instances of linking /r/ in the appropriate phonetic contexts (see § 5.5). Now regarding the lateral approximant, while English has only one such phoneme with two main allophones (clear and dark l), Spanish has two lateral phonemes: the alveolar one [l] (la, el [la] [el] ‘the’) and the palatal one [ʎ] (calle [ˈkaʎe] ‘street’), although the latter is increasingly delateralised resulting in a merger with the central palatal approximant [j] ([ˈkaje] or as a voiced palatal plosive [ɟ] word-initially (llamo ‘I call’ [’ɟamo]), so that for many speakers the following pairs are homophones: halla ‘have’ (pres subj) – haya ‘beech’, olla ‘pot’ – hoya ‘hole’, malla ‘net’ – maya ‘mayan’, pollo ‘chicken’ – poyo ‘stone bench’. Since in Spanish [l] is clear in all positions, when speaking in English SSLE should be careful to pronounce this sound only in syllable-initial positions before vowels or /j/, producing the dark variant elsewhere. To articulate dark [ɫ]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

222

Consonants

the tip of the tongue should not be curled backwards, the most essential feature being a slight rounding of the lips of the [o] or [ɔ̝] quality, this sound being similar to the l used in Catalan in such words as lámina ‘sheet’, ‘slice’ [ˈɫamina] or sal ‘salt’ [saɫ] (Finch and Ortiz Lira 1982: 64; Estebas Vilaplana 2009: 88–96; Mott 2011: 258–259; Cruttenden 2014: 222). Further practice AI 4.21 The spellings of RP /r/. Further practice right /raɪt/ bright /braɪt/ staring /ˈsteərɪŋ/

starring /ˈstɑːrɪŋ/ write /ˈraɪt/ rhyme /raɪm/

train /treɪn/ tremble /ˈtrembl ̩/

These recent songs are really surprising. /ðiːz ˈriːsnt sɒŋz ə ˈrɪəli səˈpraɪzɪŋ/ Who wrote the lyrics? /huː rəʊt ðə ˈlɪrɪks|/ Roberta ran rings around the Roman ruins. /rəˈbɜːtə ræn rɪŋz əˈraʊnd ðə ˈrəʊmən ˈruːɪnz/

AI 4.22 The spellings of RP /l/. Further practice lamp /læmp/ eleven /ɪˈlevn̩/ tall /tɔːl/

lake /leɪk/ already /ɔːlˈredi/ feel /fiːl/

lent /lent/ fill /fɪl/ lent /lent/

Lots of lakes are splendid for sailing. /lɒts əv leɪks ə ˈsplendɪd fə ˈseɪlɪŋ/ Almost all animals are still beautiful when they are old. /ˈɔːlməʊst ɔːl ˈænɪml ̩z ə stɪl ˈbjuːtəfl ̩ wen ˈðeɪ ər əʊld/ One of the loveliest flowers is called lily of the valley. /wʌn əv ðə ˈlʌvlɪɪst ˈflaʊəz ɪz kɔːld ˈlɪli əv ðə ˈvæli/ Lily ladles little Letty’s lentil soup. /ˈlɪli ˈleɪdl ̩z ˈlɪtl ̩ ˈletiɪz ˈlentl ̩ suːp/ 4.2.5.2 Glide consonants /j w/ In RP there are two /j w/ glide consonants, also known as semi-vowels or semi-consonants (see fn 13). This means that both phonemes are phonologically like consonants but phonetically like vowels (see also Gimson 1984:

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

223

211–216; Roach 2005: 63–64; Collins and Mees 2003: 172–176; Collins and Mees 2009: 92–94; Cruttenden 2014: 228–235). Phonologically, /j/ and /w/ behave like consonants in that they can only occur in the margins (onset positions) of syllables, either individually or as part of a consonant cluster. Their consonantal nature is also reinforced by the fact that articles adopt their consonantal forms before /j/ or /w/ (the university [ðə ˌjuːnɪˈvɜːsɪti], a woman [ə ˈwʊmən]), and it will be shown that their allophones are affected by the same phenomena as those undergone by other consonants. From a phonetic point of view, however, /j/ and /w/ are vowel-like in that they involve a rapid movement or glide from a position similar to [iː] (with spread or neutral lips) or [uː] (with rounded lips) respectively on to a vowel of a longer duration, the actual point of departure being affected by the nature of the following sound. IPA Symbols j Lower-case J. w Lower-case W. Identification /j/ lenis voiced palatal approximant. /w/ lenis voiced labial-velar approximant. Allophones /j/ [j]̥ ‘devoiced fricative’ [ç] ‘voiceless palatal fricative’ [tʃ ] [dʒ] ‘affricated’ [ʃ ] [ʒ] ‘fricativised’

/w/ [w̥ ] ‘devoiced’ [ʍ] ‘voiceless labial-velar fricative’

Description For the articulation of RP /j/ the soft palate is in its raised position shutting off the nasal resonator, the vocal folds vibrate, and the front of the tongue is raised towards the hard palate to approximately the position for an [i] or [ɪ]-type vowel. This is a rapid vowel-like glide on to a vowel of longer duration and prominence, its starting point being of a closer or more open variety according to the degree of openness of the following vowel. When it occurs in consonant clusters, the consonant preceding /j/ will have a palatalised articulation with some degree of lip-spreading (argue [ˈɑːɡʲjuː]. Like /j/, /w/ produces a crescendo glide that varies its starting point according to the degree of closeness of the following vowel that has greater prominence. The soft palate is also in its raised position shutting off the nasal resonator and

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

224

Consonants

the vocal folds vibrate. But in this case the tongue is raised towards the soft palate, to approximately the position for an [u] or [ʊ]-type vowel, and this glide is accompanied by strong lip-rounding. For this reason [w] is said to have a double articulation as it involves a combination of two strictures of equal rank: labial and velar open approximation. When it occurs in consonant clusters, the consonant preceding /w/ will be noticeably labialised undergoing strong lip-rounding (language [ˈlæŋɡʷwɪdʒ]. See the animations and videos under RP /w/ and /j/ (compared with the Spanish sequences /wa we wo wi/ and /ja je jo ju/ respectively) in the Sound Bank of the EPSS Multimedia Lab. Environment and main allophonic realisations Both /j/ and /w/ occur only in pre-vocalic positions at the beginning or in the middle of words, but never at the end of a word. /j/ has a devoiced realisation after voiceless consonants [j],̥ but there exists an additional fricative component when /j/ occurs after /p t k/ initially in stressed syllables (pure [pjʊə], tumult ̥ [ˈtjuːmʌlt]. It may also be realised as a completely devoiced palatal fricative ̥ [ç] similar to the German ich-Laut, usually in the sequence /h/ + [j] in a restricted number of words (hew [çuː], huge [çuːdʒ], human [ˈçuːmən], humour [ˈçuːmə]), although on the phonetic pattern /p t k/ + /j/ there are fluctuations between [hj] and [ç] pronunciations. The sequences /tj/, /dj/, /sj/ and /zj/ are normally coalesced into [tʃ ], [dʒ], [ ʃ ] and [ʒ] in word-medial syllables and at close-knit word-boundaries (educated [edʒʊˈkeɪtɪd], would you [ˈwʊdʒu], assume [əˈʃuːm], presume [prɪˈʒuːm]), but they may also be heard in stressed onset positions (tune [tʃuːn], dune [dʒuːn]). These pronunciations, however, are not always accepted in Conservative RP and may be stigmatised as “lazy speech” (see also § 5.3 for details on assimilation). A junctural [ʲ] may be pronounced to reinforce the boundary between syllableor word-final [i: i eɪ aɪ ɔɪ] and a following vowel (seeing [ˈsiːʲɪŋ], sighing [ˈsaɪʲɪŋ], my ear [maɪ ˈʲɪə]). A glottal stop may also be heard as an alternative to junctual [ʲ] (my ear [maɪ ˈʔɪə]). Now considering /w/, it has like /j/ a partially devoiced allophone [w̥ ] when it follows a voiceless consonant. But when it follows /t k/ in stressed syllableinitial positions, the devoicing is complete so that a voiceless labial-velar fricative [ʍ] is produced (twin [tʍɪn], question [ˈkʍestʃən], conquest [ˈkɒŋkʍest]). A junctual [ʷ] may be pronounced to reinforce the boundary between syllableor word-final [u: u əʊ aʊ ɔɪ] and a following vowel (doing [ˈduːʷɪŋ], no asking [nəʊ ˈʷɑːskɪŋ]), which may be replaced by a glottal stop (no asking [nəʊ ˈʔɑːskɪŋ]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

225

Spellings Table 31: The spellings of RP /j w/ /j/

/w/

yes /jes/, lawyer /ˈlɔːjə/

win /wɪn/

piano /pɪˈænəʊ/, onion /ˈʌnjən/

+ silent when /wen/, why /waɪ/

(as part of /juː/ or /jʊə/) abuse /əˈbjuːs/, endure /ɪnˈdjʊə/ nuisance /ˈnjuːsns/

(usu. /kw/) quite /kwaɪt/, quiet /ˈkwaɪət/ language /ˈlæŋɡwɪdʒ/

(as part of /juː/) (usu. Fr loanwords) argue /ˈɑːɡjuː/, residue /ˈrezɪdjuː/ barbecue /ˈbɑːbɪkjuː/ / (as part of /juː/) (usu. Fr loanwords) nephew /ˈnefjuː/, stew /stjuː/ adieu /əˈdjuː/, feud /fjuːd/ Rare cases

beauty /ˈbjuːti/ azalea /əˈzeɪliə/, but also /əˈzeɪljə/

As shown in Table 31, RP /j/ is very common as part of the sequences /juː/ and /jʊə/. The latter may be pronounced as [jɔː] (see § 3.3.2). The following fluctuations between [juː] and [uː] pronunciations should also be borne in mind (Jones 1975: 209–210; Gimson 1984: 192; Cruttenden 2014: 230). After /tʃ dʒ r/ [uː] pronunciations tend to be preferred, but [juː] is retained after plosives, /f/, /v/, and /h/ (queue /kjuː/, few /fjuː/, view /vjuː/, huge /hjuːdʒ/), as well as when /l/ is preceded by an accented vowel (value /ˈvæljuː/, curlew /ˈkɜːljuː/). Elsewhere both /juː/ and /uː/ and may be heard, although /uː/ grows increasingly common (and is the NP in AE) after /n/, /l/ and /s/ in accented syllables (neutral /ˈnuːtrəl/, lute /luːt/), whilst /juː/ remains predominant after /s z θ l/ (consume /kənˈsjuːm/, enthusiasm /ɪnˈθjuːzɪæzəm/). Note that the letter is silent in word-initial , spellings (sword /sɔːd/, wrist /rɪst/). Regional and social variants In East Anglia /j/ may be dropped after all consonants, whereas in GA it is not pronounced following /t d θ ð n/ and so such words as tune, dune or duty

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

226

Consonants

will have [uː] pronunciations. In Irish English yod-dropping also occurs in stressed syllables after alveolar sonorants (new [nuː]). In certain areas of Wales the [h] + [j] sequence is reduced to [j] (Wells 1982: 436; Hickey 2007: 307). The most remarkable variant of /w/ concerns the spelling . It is pronounced as a voiceless labial-velar fricative [ʍ] mainly in American English, SSE and Irish, thereby marking a contrast between wear /weə/ and where /ʍeə/, which [ʍɪtʃ ] and witch [wɪtʃ ]). [ʍ] may also be heard in Conservative RP and in formal declamatory styles (Wells 1982: 229, 432; Hickey 2007: 319). Comparison with Spanish and advice Most descriptions of the Spanish phonemic system do not include [j] and [w] as these two sounds are considered as allophones of /i/ (nieve ‘snow’ [ˈnjeβe]) and /u/ (huevo ‘egg’ [ˈweβo]) (Alarcos Llorach 1961; Navarro Tomás 1966 [1946] Quilis and Fernández 1982). Although the articulation of these two sounds should involve no problem, SSLE may find it problematic to pronounce them correctly because of the different phonetic status that they have in English and Spanish, and due to the influence of orthographic pronunciation. Since in Spanish words that begin with may be pronounced with as a voiced palatal fricative [ ʝ], a voiced palato-alveolar affricate [dʒ], a voiced palatal lateral [ʎ] or even a voiced palatal plosive [ɟ] (word-initially) (ya ‘already’ [ʝa], [dʒa], [ʎa], [ɟa]), it is common for SSLE mispronounce RP /j/ as [ ʝ], [dʒ] or even [ʎ] or [ɟa]), so that such words as jam /dʒæm/ and yam /jæm/ are often confused. SSLE should therefore remember that the spelling in English corresponds to /j/, not to [ ʝ], [dʒ], [ʎ] or [ɟ]. RP /w/, on the other hand, may be mispronounced as a [b], or quite frequently, [g] or [ɣ] may be inserted before the sound itself, so that English words like when [wen] or whisky [ˈwɪski] may be mispronounced as *[ˈɡwen] and *[ˈɡwɪski]. As in RP no sound is inserted before word-initial [w], SSLE should try to avoid the insertion of an initial [ɡ] in this context. Particularly troublesome for SSLE are the English sequences [ji], [jɪ], [wu], [wʊ]. The reason is that in Spanish, [j] and [w] may occur before all vowels except [i] and [u], whereas in English /j w/ may be found before all vowels, including /i, ɪ, u, ʊ/. For this reason, SSLE tend to wrongly pronounce words such as year /ˈjiə/ as *[iər], *[ ʝiər], *[dʒiər] or *[ʎiər] and woman /ˈwʊmən/ with initial [gu-], [ɣu-], [u-]. To pronounce RP [j] correctly, the front of the tongue must be raised to the closest variety of [i] without producing friction, and then glide quickly to a more open quality. In the case of RP [w] articulation should begin with strong lip-rounding and with the tongue being raised as for a very close variety of [u]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

227

without producing friction, and then a glide should be quickly made to a more open articulation. Moreover, as both [j w] can occur word-initially, SSLE should refrain from producing an epenthetic consonant before them (Mott 2011: 131; Sedláčková 2010: 59; Estebas Vilaplana 2009: 71–73, 86–87; Pennock 2001; Cuenca-Villarín 1996; Finch and Ortiz Lira 1982: 64, 71). Further practice AI 4.23 The spellings of RP /j/. Further practice yellow /ˈjeləʊ/ few /fjuː/ piano /pɪˈænəʊ/

yes /jes/ use /ˈjuːs/ new /njuː/

yesterday /ˈjestədi/ unusual /ʌnˈjuːʒʊəl/ valuable /ˈvæljʊəbl ̩/

Are you used to using online newspapers? /ə ju ˈjuːst tə ˈjuːzɪŋ ˈɒnˌlaɪn ˈnjuːspeɪpəz/ Is your friend a lawyer yet? Yes, yesterday he was given a job for a year in New York. /ɪz jə frend ə ˈlɔːjə jet | jes | ˈjestədi hi wəz ɡɪvn̩ ə dʒɒb ˈfɔːr ə jɜːr ɪn njuː jɔːk/

AI 4.24 The spellings of RP /w/. Further practice what /ˈwɒt/ quick /kwɪk/ when /wen/

window /ˈwɪndəʊ/ while /waɪl/ quite /kwaɪt/

question /ˈkwestʃən/ woman /ˈwʊmən/

This wet weather worries me, I wonder when it will stop. /ðɪs wet ˈweðə ˈwʌrɪz miː | ˈaɪ ˈwʌndə wen ɪt wɪl stɒp/ The first sound in ‘wood’, ‘woman’, and ‘wool’ is the same as in ‘wed’, ‘women’ and ‘well’. /ðə fɜːst ˈsaʊnd ɪn wʊd | ˈwʊmən | ənd wʊl z ðə seɪm əz ɪn wed | ˈwɪmɪn ənd wel/ Wise whales keep away from the windy west coast. /waɪz weɪlz kiːp əˈweɪ frəm ðə ˈwɪ ndi westkəʊst/ While we were walking, we were watching window washers wash Washington’s windows with warm washing water. /waɪl wi wə ˈwɔːkɪŋ | wi wə ˈwɒtʃɪŋ ˈwɪndəʊ ˈwɒʃəz wɒʃ ˈwɑːʃɪŋtənz ˈwɪndəʊz wɪð wɔːm wɒʃ ˈwɔːtə/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

228

Consonants

Further reading As in the previous chapter, we recommend Cruttenden (2014, Chapter 9), Roach (2005, Chapters 4, 6 and 7), Collins and Mees (2003: 139–188) and Gimson (1984, Chapter 8) for an examination of English consonants. More specialised and technical analyses may be found in Ashby (2005, Chapters 9 and 10), Ashby and Maidment (2005, Chapters 6 and 8), Ladefoged (2005, Chapters 11 and 14,) and Ladefoged and Maddieson (1996, Chapter 10) and Laver (1994, Chapters 11 to 14). A more advanced account of the acoustic features of consonants is offered by Hayward (2000, Chapter 7). For more thorough descriptions of consonantal phonemes in Spanish, the reader is again referred to Martínez Celdrán and Fernández Planas (2007), Navarro Tomás (1991[1918], 1966 [1946]), Quilis and Fernández (1982), Quilis (1985, 1993), and Alarcos LLorach (1961 [1983]), as well as to Martínez Celdrán and Fernández Planas (2003), Colina (2009) and Gil (2007), to mention but a few basic references. To learn more about the acoustics of Spanish consonants, good references to consult are Cabrera Abreu and Vizcaíno Ortega (2009), Martínez Celdrán (1998) and Quilis (1981). Lastly, additional accounts of RP in contrast with those of Spanish or addressed to SSLE may be found in the aforementioned (mostly) well-established studies: Estebas Vilaplana (2009, Chapters 2 and 3), Monroy Casas (1980, 1981, 2012, Chapters 1 and 2), Mott (2011, Chapters 5 and 12), Núñez Méndez (2005), Finch and Ortiz Lira (1982, Chapter 7), Sánchez Benedicto (1980) and Stockwell and Bowen (1965); whereas issues of acquisition of English consonants by speakers of Spanish are discussed in Gallardo del Puerto (2005), Dale and Poms (1986), Coe-Guerrero (1981), among many others.

Exercises 1. Passive transcription. Read aloud the following passage given in broad transcription, paying special attention to the consonant sounds. Then write it out in ordinary spelling. / fənɒˈlɒdʒɪkli / ˈsɪləbl ̩z kən bi lʊkt ət ˈteɪkɪŋ ˈɪntə əˈkaʊnt ðə ˈpɒsəbl ˌkɒmbɪ ˈneɪʃnz əv ˈfəʊniːmz ðət ər əˈlaʊd ɪn ə ɡɪvn̩ ˈlæŋɡwɪdʒ // sɪˈlæbɪk ˈstrʌktʃə ˈdɪfəz frəm wʌn ˈlæŋɡwɪdʒ tu əˈnʌðər ɪn tɜːmz əv sʌtʃ pəˈræmɪtəz əz ˈhaʊ meni ˈkɒnsənənts kən əˈkɜːr ɪn ði ˈɒnsets ənd ˈkəʊdəz əv ˈsɪləbl ̩z / ɔː ˈweðə ˈvaʊəlz kən əˈkɜː wɪðˈaʊt ˌkɒnsəˈnæntl ̩ ˈɒnsets / ˈsɪləbl ̩z wɪð ˈzɪərəʊ ˈɒnsets / ɔː ˈweðə bəʊθ ˈəʊpn̩ / ˈendɪŋ ɪn ə ˈvaʊəl / ənd kləʊzd / ˈendɪŋ ɪn ə ˈkɒnsənənt / ˈsɪləbl ̩z ə ˈpɒsəbl ̩ //

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

229

A comparison of English and Spanish consonants

2. Provide words that have the specified fricative and affricate sounds in initial, medial and final the positions: /f, v, θ, ð, s, z, ʃ, ʒ, ʧ, ʤ, h/ 3. Stops: /p, b, t, d, k, g, m, n, ŋ/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /ˈgɑːdn̩ / 6. /ˈræbɪt/ 11. /ˈsɪŋə/ 16. /sʌŋ/

2. /æd/ 7. /ˈsɪnə/ 12. /wɒt/ 17. /ˈbæŋə/

3. /ˈbegɪŋ/ 8. /ˈbʌtə/ 13. /mɒb/ 18. /pɑːm/

4. /ˈkɒfɪŋ/ 9. /ˈkɒfɪn/ 14. /ˈnɔːti/ 19. /sɪŋk/

5. /ˈtempə/ 10. /bæg/ 15. /ˈmeɪkɪŋ/ 20. /gæp/

4. Fricatives: /f, v, θ, ᵭ, s, z, ʃ, ʒ, h/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /briːf/ 6. /ədˈvaɪz/ 11. /ˈzenɪθ/ 16. /ˈhɪsɪŋ/

2. /θaɪ/ 7. /ˈpleʒə/ 12. /briːv/ 17. /ˈmeθəd/

3. /ədˈvaɪs/ 8. /ˈmʌðə/ 13. /ˈmæʃɪŋ/ 18. /bɪˈheɪv/

4. /ˈvɜːʒn̩/ 9. /ˈiːzi/ 14. /ðeə/ 19. /feɪs/

5. /suːð/ 10. /ruːʒ/ 15. /ˈʃerɪ/ 20. /ˈsʌfə/

5. Affricates and Approximants: /ʧ, ʤ, l, r, j, w/. Read aloud the following words given in broad transcription and then write them out in ordinary spelling. 1. /bæʧ/ 6. /njuː/ 11. /eʤ/ 16. /ʤeɪn/

2. /kwiːn/ 7. /rɪəl/ 12. /eʧ/ 17. /ˈkwaɪət/

3. /weɪz/ 8. /jes/ 13. /ˈwɒʧɪŋ/ 18. /juːs/

4. /bæʤ/ 9. /ˈlɒri/ 14. /ʤæm/ 19. /ʧɜːʧ/

5. /ˈrɪəli/ 10. /ʧɑːʤ/ 15. /jɔːnd/ 20. /ʤuːs/

6. Mixed consonant sounds. Identify the following words from their transcription. Read them aloud first, and then write the spelling. 1. /ʧeɪn/ 6. /kəʊld/ 11. /ˈlɪsənɪŋ/ 16. /vjuː/ 21. /ˈfjuːʧə/ 26. /ˈstrʌgl ̩/

2. /ɜːʤ/ 7. /ˈkɒlə/ 12. /sniːz/ 17. /ˈpɪkʧə/ 22. /ˈpʌzl ̩d/ 27. /ˈbʌtn̩/

3. /ˈrɪŋɪŋ/ 8. /ʤʌg/ 13. /ˈplʌmə/ 18. /ˈhʌzbənd/ 23. /liːg/ 28. /streɪt/

4. /ˈrɪbn̩/ 9. /ˈnɪəlɪ/ 14. /sæŋ/ 19. /əˈraɪvl ̩/ 24. /mʌʧ/ 29. /ˈbeɪkn̩/

5. /fjuː/ 10. /ˈiːvnɪŋ/ 15. /ˈθaʊzənd/ 20. /ˈbriːðɪŋ/ 25. /ˈleɪzi/ 30. /ʤeɪn/

7. /t, θ, ð/. Classify the words given below in spelling into three columns. Note that there is word which does not fit in any column. eighth third though

method the soften

then this through

outhouse time brother

both three eight

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

230

Consonants

8. /n/-/ŋ/. Classify the words given below in spelling into two columns. Note that there are two words which do not fit in either column. long funny angry

alone sink fin

autumn sinner hymn

ginger singer danger

finger anchor clingy

9. /s, z, ʃ, ʒ, ʧ, ʤ/. Classify the words given below in spelling into six columns. Note that there are two words which do not fit in any column. school sugar fashion gym nature picture

general equation journal beds machine muscle

miss Japan science leisure Celtic vision

dessert aisle scene prison treasure bets

raise cello accept chore jelly mission

10. Transcription. Transcribe the following words, paying special attention to fricative and affricate sounds (broad transcription). these season breath witch rush

fifth measure lose offering rouge

behave lashes wish loose match

scissors pleasure thin breathe laugh

feathers mash lazy over gem

11. Transcribe the following sentences. 1. The grassy fields are closed to the cows 2. The face of her niece looks satisfied in the picture. 3. Zoom in a thousand times! 4. The famous Irish mansion near the station hides a shocking treasure. 5. There was confusion and division after the ‘Prestige’ issue. 6. My brother bought our mother this smooth leather jacket. 7. She thanked both authors with enthusiasm and a thoughtful smile. 8. You have a very good view of the valley and the river. 9. Photographs of African elephants for sale. 10. Whose behaviour is honest and honourable in his household? 12. Homophones. The following transcriptions represent a pair of homophones (unless otherwise stated). Give their spellings. 1./ˈmɑːʃl ̩/ 5. /hɜːd/ 9. /ʃɪə/

2. /ˈmedl ̩/ 6. /reɪn/ (3) 10. /reɪz/

3. /raɪt/ (4) 7. /ˈneɪvl ̩/ 11. /sʌm/

4. /rɪŋ/ 8. /ˈmʌsl ̩/ 12. /ˈsteɪʃnri/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

A comparison of English and Spanish consonants

231

13. A contrastive exercise. Pronounce the following words given in spelling as if they were English words first, and then Spanish. Explain what makes them different in terms of the initial consonant: pan

ten

kilo

red

rural

van

14. Contrastive exercise. Aspiration. Try and pronounce the following English words, paying special attention to the /p, t, k/ phonemes: palm

pool

train

top

class

cot

Now, pronounce the following English words, in which the plosives are in a different position: spam

spell

strain

stop

scatter

school

Can you tell the difference between the plosives in both series? Now, pronounce the following similar Spanish words: palma pulpo tren tope clase cota. Which of the two series above do these resemble more? Finally, try and swap the English aspirated plosives by the Spanish counterparts, and viceversa, achieving a foreign effect. 15. English /h/ versus Spanish /x/. Pronounce both sounds in turn, going from one to the other – i.e. from glottal to velar and back again: /h/-/x/-/h//x/-/h/-/x/. Now, pronounce the English word ‘ham’ and the Spanish word ‘jamón’; now interchanging the initial phonemes so that each word is pronounced as if in the ‘fashion’ of the other language – i.e. with a foreign accent. On the basis of this activity, explain what makes someone sound foreign. 16. There are certain consonant sounds in English which are not part of the Spanish repertoire, even if the actual letter does exist. This is the case of /b/ and /v/, for example. Pronounce the following words carefully, paying attention to the sounds /b/ and /v/, in different positions in the word. Give their transcription. bat beer berry best burbs (short for suburb) lobes rebel dribble curb dub

vat veer very vest verbs loaves revel drivel curve dove

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

232

Consonants

17. The /s/ – /ʃ/ contrast that exists in English does not occur in Spanish. Proceed in the same way as in exercise 16. save see seal seat single seen sock sew crust gust crass diss gas plus plus puss

shave she she’ll sheet shingle sheen shock show crushed gushed crashed dished gashed plushed plushed pushed

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:09 PM

Chapter 5

5 Segment Dynamics: Aspects of Connected Speech 5.1 Introduction In this chapter we are not going to look at sounds in isolation, but rather our attention is centred on connected sequences of sounds in words and larger units. The term connected speech refers to an “utterance consisting of more than one word” (Gimson 1984: 255). One of the distinctive features of connected speech is that when uttered in a chain speech sounds affect each other and their features change in such a way that allophonic variations, or different realisations of phonemes, occur as a result of the operation of a number of (allophonic) rules which, in general, can be described as simplification rules. Speakers often have to find a balance between two conflicting factors: ease of articulation and distinctiveness (Boersma 1998). Listeners expect to hear words as different and distinct from one another as possible, but, at the same time, speakers generally prefer to reduce articulatory effort whenever possible, and so they tend to resort to simplification mechanisms so as to convey only the phonetic information that is felt necessary. Although there can be a great deal of variability in how words are pronounced and there are many factors at work in such process (Bell et al., 2003; Shockey 2003), listeners rarely experience difficulty in understanding. What follows describes two different kinds of connected speech phenomena that are involved in phonetic variation: (1) those affecting sounds within words and/or at word (syllable and morpheme) boundaries, as in the case of coarticulation (Section 5.2.), assimilations (Section 5.3), elision (Section 5.4), linking (Section 5.5) and juncture (Section 5.6). (2) those affecting the word as a whole, which is the case of gradation (including neutralisations of weak forms) (Section 5.7), and variations in accentual patterns, the latter being illustrated in Chapter 6 together with other prosodic features of speech.

5.2 Co-articulation and allophonic variations of vowels and consonants When we speak, as already explained, sounds are uttered in sequence (one after the other) in an articulatory continuum. As a result, features of one segment

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

234

Segment Dynamics: Aspects of Connected Speech

(or sound) may be found in an adjacent segment. When this occurs the label coarticulation is generally used to refer to “patterns of coordination, between the articulatory gestures of neighboring segments, which result in the vocal tract responding at any one time to commands for more than one segment” (Manuel 1999: 179; Crystal 2008: 82). Another label that is employed to describe this phenomenon is coproduction to emphasise that there exists temporal cooccurrence or overlap in the articulation of two (or more) gestures (Byrd 1996: 210). Examples of coarticulation are the production of labialised (cool [kʷuːɫ]) or palatalised (key [kʲiː]) allophones, which are produced when consonants are in contact with rounded vowels and front spread vowels, respectively. These types of allophonic alternations may be found within the word, and also at word boundaries, as shown in Table 32 that summarises the main allophonic realisations of consonants (Gimson 1984: 283–298; Cruttenden 2014: 270–322). For an exhaustive description of the allophonic variations of each RP vowel and consonant, please refer to the relevant (sub)sections in Chapters 3 and 4. 5.2.1 Lengthening and clipping In English long (pure) vowels and vowel glides (diphthongs and diphthongs plus schwa) are equivalent in length and are subject to the same variations of quantity. All undergo pre-fortis clipping (Denes 1955; Wiik 1965: 114; Monroy Casas 1980: 75) and therefore are reduced in length if they precede a fortis consonant (e.g. place [pleˑɪs] vs. plays [pleːɪz]). The reduced forms show a considerable shortening of the vowel or the first element of the glide (half-long [ˑ]) (Petterson and Lehiste 1960). As already remarked, not only vowels but also any sonorant consonant is vulnerable to pre-fortis clipping before a fortis consonant. So comparing sent [senˑt] and self [seɫˑf] with send [senːd], selves [seɫːvz], the /en/ and /el/ sequences are shorter in the first pair of words as a result of prefortis clipping. In addition, length reduction may also derive from so-called rhythmic clipping, which occurs when the duration of a stressed vowel is reduced as a result of the addition of one or several syllables to the same foot.83 Compare the durations of the stressed vowel in read, reader and readership, where the stressed syllable has progressively less duration as progressively more syllables follow within the same stress group, so that the first instance of [iː] in read is longer than the second [iˑ] in reader, and this, in turn, is longer than the third [iˑ] in readership (Fudge 1984: 20). 83 Crystal (2008: 193) defines the term “foot” as “the unit of rhythm in languages displaying isochrony, i.e. where the stressed syllables fall at approximately regular intervals throughout an utterance.”

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Co-articulation and allophonic variations of vowels and consonants

235

Table 32: Main allophonic variations of consonants Type place of articulation

Within the word

At word boundary

/t/: post-alveolar in tray [t ̠ɹ ̥eɪ]

/t/: dental in not those [ˈnɒt ̪ðəʊz],

/k/: advanced in king [k̟ɪŋ̟]

/d/: dental in hide this [ˈhaɪd̪ðɪs]

/n/: dental in ninth [naɪn̪θ]

/m n/: labio-dental in men fight [meɱ faɪt]

/m n/: labio-dental in comfort [ˈkʌɱfət]

/s/: retracted in horse-riding [hɔː s̠ˈɹaɪdɪŋ] voice

/r/: devoiced in upright [ˈʌpɹ̥aɪt],

/r/: devoiced in at rest [әtˈɹ̥est]

/w/: completlety devoiced in twin [tʍɪn]

/w/: devoiced in at once [әtˈw̥ ʌns]

/l/: devoiced in climb [kl ̥aɪm]

/l/: devoiced in at last [әtˈl ̥ɑːst]

/j/: devoiced in pure [pjʊə] ̥

/j/: devoiced in thank you [ˈθæŋkjuː] ̥

/m n ŋ/ slightly devoiced in smart [sm̥ ɑːt] st], sneeze [sn̥iːz], bacon [ˈbeɪkŋ̊] lip position

lip-spreading /p/ /t/ /k/ /m/ /n/ /ŋ/ /l/ /r/ /f/ /s/ /h/ /ʃ/

nasal resonance

lip-rounding

lip-rounding

pea [pʲiː] tea [tʲiː] keep [kʲiːp] mule [mʲjuːl] noon [nʲuːn]

pool [pʷuːl] too [tʷuː] cool [kʷuːɫ] mob [mʷɒb] no [nʷəʊ]

at one [ətʷˈwʌn] pack one [ˈpækʷwʌn] come once [ˈkʌmʷwns]

lea [lʲiː] read [ɹʲiːd] feel [fʲiːl] see [sʲiː] he [hʲiː] sheer [ ʃʲɪə]

loo [lʷuː] room [ɹʷuːm] fool [fʷuːɫ] soon [sʷuːn] who [hʷuː] sugar [ˈʃ ʷʊɡə]

Nasalisation of vowels preceding /m n/ in ham [hæ̃ m] Nasalisation of vowels surrounded by or surrounding nasals: man [mæ̃ n] Nasalisation of /l/ before or after ̃ nasals: signalman [ˈsɪɡnɫmən]

long one [ˈlɒŋʷwʌn] will we [ˈwɪɫʷwiː]

this way [ðɪsʷˈweɪ]

Nasalisation of vowels when there is a nasal in the adjacent word, come along [kʌ̃m ə̃ˈlɒŋ], especially if another adjacent nasal occurs in the word containing the vowel: bring in [brɪ ̃ŋ ɪ ̃n] Nasalisation of approximants: tell ̃ me [ˈteɫmiː]

The clipping-effect is more noticeable if reduced realisations are contrasted with those occurring in lengthening environments (i.e. open word-final syllables), where long (pure) vowels and vowel glides may have extra-long realisations [ːː] (play [pleːːɪ]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

236

Segment Dynamics: Aspects of Connected Speech

AI 5.1 Allophonic variation of vowels Below are listed more examples of (1) extra-long [ːː], (2) normally-long [ː], and (3) reduced [ˑ] allophonic realisations of English vowels and glides: [iː]

[ɜː]

[iːː]: tea [tiːː], see [siː] [iː]: team [tiːm], seed [siːd] [iˑ]: meet [miˑt], leak [liˑk]

Compare: see [siːː] seed [siːd]

[ɜːː]: purr [pɜːː], myrrh [mɜːː] [ɜː]: girl [gɜːl], word [wɜːd] [ɜˑ]: skirt [skɜˑt], work [wɜˑk]

Compare: purr [pɜːː]

seat [siˑt]

purl [pɜːl]

pert [pɜˑt]

[ɑː] [ɑːː]: car [kɑːː], star [stɑːː] [ɑː]: calm [kɑːm], starred [stɑːd] [ɑˑ]: cart [kɑˑt], start [stɑˑt]

Compare: are [ɑːː] aren’t [ɑːnt]

[ɔː]

Compare: core [kɔːː] cored [kɔːd] caught [kɔˑt]

[ɔːː]: poor [pɔːː], saw [sɔːː] [ɔː]: poured [pɔːd], sawed [sɔːd] [ɔˑ]: port [pɔˑt], sought [sɔˑt]

art [ɑˑt]

[uː] [uːː]: blue [bluːː], two [tuːː] [uː]: pool [puːl], food [fuːd] [uˑ]: fruit [fruˑt], Luke [luˑk]

Compare: shoe [ ʃuːː] shoed [ ʃuːd] shoot [ ʃuˑt]

[eɪ] [eːːɪ]: day [deːːɪ], bay [beːːɪ] [eːɪ]: fade [feːɪd], game [geːɪm] [eˑɪ]: eight [eˑɪt], face [feˑɪs]

Compare: say [seːːɪ] save [seːɪv] safe [seˑɪ f]

[eә] [eːːә] there [ðeːːә], pear [peːːә] [eːә] chairs [ʧeːәz], cared [keːәd] [eˑә] scarce [skeˑәs]

Compare: scare [skeːːә] scares [skeːә] scarce [skeˑәs]

[aɪ] [aːːɪ]: fly [flaːːɪ], die [daːːɪ] [aːɪ]: mine [maːɪn], hide [haːɪd] [aˑɪ]: fight [faˑɪt], like [laˑɪk]

Compare: tie [taːːɪ] time [taːɪm]

[aʊ] [aːːʊ]: how [haːːʊ], cow [kaːːʊ] [aːʊ]: town [taːʊn], loud [laːʊd] [aˑʊ]: shout [ ʃaˑʊt], mouse [maˑʊs]

Compare: allow [әˈlaːːʊ] allows [әˈlaːʊz] a louse [ә laˑʊs]

[әʊ] [әːːʊ]: go [gәːːʊ], toe [tәːːʊ] [әːʊ]: home [hәːʊm], road [rәːʊd] [әˑʊ]: goat [gәˑʊt], both [bәˑʊθ]

Compare: row [rәːːʊ] robe [rәːʊb]

rope [rәˑʊp]

[ɔɪ] [bɔːːɪ]: boy [bɔːːɪ], toy [tɔːːɪ] [bɔːɪ]: void [vɔːɪd], coin [kɔːɪn] [bɔˑɪ]: choice [ʧɔˑɪs]

Compare: joy [ʤɔːːɪ] join [ʤɔːɪn]

joist [ʤɔˑɪst]

tight [taˑɪt]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Co-articulation and allophonic variations of vowels and consonants

[ɪә] [ɪːːә]: dear [dɪːːә], here [hɪːːә] [ɪːә]: beard [bɪːәd], weird [wɪːәd] [ɪˑә]: pierce [pɪˑәs], fierce [fɪˑәs]

Compare: fear [fɪːːә]

fears [fɪːәz]

237

fierce [fɪˑәs]

[ʊә] [ʊːːә]: poor [pʊːːә], moor [mʊːːә]84 Compare: pour [pʊːːә] poured [pʊːәd] [ʊːә]: cured [kjʊːәd], moored [mʊːәd] [ʊˑә] (rhythmic clipping): security [sɪˈkjʊˑərɪti]

5.2.2 Voicing and devoicing Voicing refers to the auditory result of the vibration of the vocal folds (Crystal 2008: 515). We have seen that voiced sounds are produced while the vocal folds are vibrating, whereas voiceless or unvoiced sounds are articulated with no such vibration. Likewise, voicing is the term applied to describe the phenomenon whereby voiceless phonemes (/p/, /t/, /k/, /ʧ/, /f/, /θ/, /s/, /ʃ/, /h/) have voiced allophonic realisations (annotated as [ˬ] or [ˇ] in intervocalic positions and between voiced sounds (e.g. matter [ˈmæt̬ər]). We have also explained that the opposite phenomenon exists in English. In particular phonetic environments (word-initially and word-finally, as well as between voiceless sounds including close-knit word boundaries), sounds which are normally voiced (/b/, /d/, /g/, /ʤ/, /v/, /ð/, /z/, /ʒ/, /m/, /n/, /ŋ/, /w/, /l/, /r/, /j/) are articulated with less vibration than elsewhere and are said to be devoiced (symbolised with an under- [˳] or over-ring [˚]) (pig [pɪɡ̊], give [gɪv̥], breathe [briːð̥], his [hɪz̥], George [d̥ʒɔːd̥ʒ]̊ ); whereas they are fully voiced if they occur between vowels (labour [ˈleibər]). Approximants are also devoiced if they are preceded by a fortis plosive in consonantal clusters, both within a word and at close-knit word boundaries. This occurs in e.g. quiet [ˈkw̥ aɪət] try [tɹ̥aɪ], play [pl ̥eɪ], pure [pȷ ̊ʊə], at last [ətl ̥ɑːst], at once [ətw̥ ʌns], thank you [θæŋkȷ ̊uː], where the aspiration of fortis plosives, whatever its degree, is manifested as devoicing of the subsequent approximants. It should be noted, however, that approximants are only slightly devoiced when they are preceded by a fortis consonant in an unstressed syllable (chaplain [ˈtʃæpl ̥ɪn]), while nasals /m n ŋ/ are also slightly devoiced if they are preceded by voiceless plosives (certain [ˈsɜːtn̩̥], smoke [sm̥ əʊk], bacon [ˈbeɪkŋ̩̊]). In addition, when clusters of lenis consonants occur in a devoicing environment, both sounds lose some degree of vocal fold vibration, but it is the one nearer to the devoicing factor that becomes completely devoiced and, for practical purposes, it is the only one that is marked as devoiced in allophonic

84 Remember, however, that poor and moor are generally pronounced with /ɔː/ (see § 3.2.4).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

238

Segment Dynamics: Aspects of Connected Speech

transcriptions. Thus, although in pigs [pɪɡz̥] both [ɡ] and [z] lose some of their voicing because they occur word-finally, it is only [z̥] that is marked as completely devoiced, as it occurs immediately before silence (Lombardy 1999; Shockey 2003; Ernestus et al. 2006).

5.2.3 Lip shape Consonants may be modified by having a secondary articulation under the influence of the lip position adopted for the articulation of adjacent vowels or semivowels (see § 2.3.2.5) (Shockey 2003; Cruttenden 2014: 308–310). Two types of allophones are derived from this process: Labialised articulations [ʷ], which involve lip-rounding with the accompanying elevation of the tongue back and occur next to rounded vowels /ɒ ɔː ʊ uː/ and the rounded semi-vowel [w] (Ladefoged and Maddieson 1996: 368). Within words, the effect is most noticeable in the consonants preceding the rounded vowel or semivowel, which explains that only these consonants tend to be marked as labialised in narrow transcriptions (bog [bʷɒɡ], bloom [blʷuːm], pork [pʷɔˑkʷ], rude [ɹʷuːdʷ], soon [sʷuːnʷ], upward [ˈʌpʷwəd], whose [hʷuːzʷ]). At word boundaries, the semi-vowel [w] exerts some degree of labialisation on the preceding consonant (that one [ˈðætʷ wʌn], wrong one [ˈɹʷɒŋʷ wʌn]), but a rounded vowel in an adjacent word does not seem to exert the same labializing influence (e.g. in Who takes this? the [uː] of who does not labialise the [t] of takes, but it does labialise the preceding [hʷ]). Palatalised articulations [ʲ], which are produced with lip-spreading and raising of the front of the tongue under the influence of adjacent spread vowels /ɪ iː/ and the palatal semivowel [j], either within words and at close-knit word boundaries (Barry 1992). As in the case of labialisations, palatalisations are more noticeable in the consonants preceding /ɪ iː j/ than in consonants following them, which explains that only the former tend to be marked in narrow transcriptions (bee [bʲiː], cute [kʲjuˑt], pea [pʲiː], fin [fʲɪnʲ], read [ɹʲiːdʲ]).

5.2.4 Nasal resonance Vowels and vocalic series have nasalised [˜] allophonic articulations, i.e. are articulated with nasal resonance by lowering the soft palate, in the vicinity of a nasal consonant (Cohn 1990). This tendency can be more noticeable in varieties of English other than RP such as GA, where nasalisation not only occurs in vowels next to nasal consonants but also in vowels which are not in contact

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Co-articulation and allophonic variations of vowels and consonants

239

with any nasal sound, which infuses some American accents with an overall nasal quality (He drives too fast [hĩ draɪ ̃vz tũː fæ̃ st]) (Estebas Vilaplana 2009: 85). For practical purposes, these will only be marked when they are particularly noticeable, that is, when occurring between nasals or when a nasal follows them, within words (man [mæ̃ n], singer [ˈsɪ ̃ŋə], ton [tʌ̃n], ham [hæ̃ m]). Somewhat reduced nasalisations occur in the same contexts between words (bring in [brɪ ̃ŋ ɪ ̃n], for me [fə̃ miː], pretty mean [ˈprɪtĩ ˈmĩːn]), but they can affect the vowel of an adjacent word that is only preceded, but not followed by a nasal (come along [kʌ̃m ə̃ˈlɒŋ]).]). Consonants can also be produced with nasal resonance. Approximants may be nasalised broadly in the same phonetic environments as vowels (helmet ̃ [ˈhelmɪt], tell mum [ˈtel ̃ ˈmʌ̃m]), while plosives have a nasal release in consonantal clusters when followed by a nasal consonant (see § 2.3.2.5, 2.4.1, 5.2.6).

5.2.5 Aspiration We have seen that in English aspiration refers to the audible breath that may accompany the articulation of fortis plosives in specific phonetic contexts (see § 4.2.1). Though varying in intensity, for teaching purposes we only recognise two degrees of the aspiration effect: aspirated [pʰ tʰ kʰ], in syllable-initial stressed position ( pan [pʰæn]), or unaspirated [p= t= k=], when preceded by [s] (ski [sk=iː]), in unaccented syllables ( police [p=əˈliːs]), and in both word and syllable final positions (tap [tʰæp=]). But if fortis plosives are followed by approximants, especially in syllable-initial stressed position, then the aspiration of the former is manifested, regardless of its degree, as approximant devoicing [l ̥ ɹ̥ ȷ ̊ w̥ ] (e.g. play [pl ̥eɪ], splendid [ˈspl ̥endɪd], try [tɹ̥aɪ], strange [stɹ̥eɪndʒ], obscure [əbˈskȷ ̊ʊə], quite [kw̥ aɪt=]). In more detailed analyses, pre-aspiration after the consonant may be distinguished from post-aspiration (aspiration after the consonant), as both features may occur in a language (e.g. Gaelic).

AI 5.2 Aspiration of voiceless plosives It has also been explained that the acoustic correlate of aspiration is the Voice Onset Time (VOT), that is, the interval that exists between the release of a fortis plosive and the voicing of the following vowel or sonorant. Broadly, aspirated voiceless plosives have positive VOTs and therefore trigger delayed onsets to voicing. In contrast, (de)voiced and unaspirated voiceless plosives, with negative and zero VOTs respectively, produce non-delayed onsets to voicing.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

240

Segment Dynamics: Aspects of Connected Speech

5.2.6 Types of release This feature is considered with regard to the plosive group only. In terms of type of release, plosives have three different allophonic realisations that stand in complementary distribution, that is, they occur in mutually exclusive environments: (1) Non audible release [p˺ t˺ k˺ b˺ d˺ ɡ˺] in consonantal clusters, if followed by another plosive or by an affricate either within a word (act [æk˺t], picture [ˈpɪk˺tʃə]). (2) Nasal release [pⁿ tⁿ kⁿ bⁿ dⁿ ɡⁿ], when the plosive is followed by a nasal consonant both between and within words (top nine [tɒpⁿ naɪn], submerge [səbᵐˈmɜːdʒ]). (3) Lateral release [pˡ tˡ kˡ bˡ dˡ ɡˡ], if the plosive is followed by a lateral consonant both within and between words (e.g. little [ˈlɪtˡɫ̩], bad leg [bædˡ leɡ]). For further details on the types of release exhibited by plosive consonants, the reader is referred to Sections 2.3.2.5 and 4.2.1.

5.2.7 Place of articulation The degree of openness or backness of a vowel and the actual place of articulation of a consonant is conditioned by its phonetic environment. Besides having closer [˔] or more open [˕] realisations, English vowels may also have more advanced [ ]̘ or more retracted [˫] realisations depending on regional variants, or otherwise on the place of articulation of the adjacent consonant which may be produced further to the front in the oral chamber (e.g. loop [lu̘ ˑp] where [u̘ ] is more fronted due to the presence of the following bilabial consonant) or further to the back (e.g. result [rɪˈzʌ̙ɫt] where [ʌ̙] results from the influence of [ɫ]). Consonants, on the other hand, may also have more advanced [ ]̟ or more retracted or backed [ ]̠ places of articulation both in consonantal clusters or in C + V series if the second element has a more advanced or a more retracted, or backed, place of articulation respectively, both within a word and at close-knit word boundaries. For practical purposes, more advanced/retracted places of articulation will only be signalled in those cases in which it is most noticeable, that is, when there exists a succession of two opposite articulatory points in the oral chamber: (1) When a velar consonant is followed by a (post)alveolar approximant, it will have a post-velar articulation ( gray [ɡ̄ ɹeɪ]), but when there is an adjacent front vowel [iː ɪ] or glide [j] (key [k̟iː]), the place of articulation is pre-velar.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Co-articulation and allophonic variations of vowels and consonants

241

(2) When a bilabial or an alveolar consonant is followed by a post-alveolar approximant in consonantal clusters (brush [b̠ɹʌʃ ], unreal [ʌn̠ˈɹɪəl], ballroom [ˈbɔˑl ̠ɹuːm], tray [t̠ɹeɪ], dry [d̠ɹaɪ]) and when they are next to a back vowel [uː ʊ ɒ ɔː] or semivowel [w] (pool [p̠ʰuːɫ], tool [t̠ʰuːɫ]), they are produced with post-bilabial and post-alveolar places of articulation, respectively. Other allophonic realisations concerning place of articulation that can be found both within words and at (close-knit) word boundaries and that should be marked in allophonic transcriptions are the following: (3) Within the nasal group: a. Labio-dental realisations [ɱ] of /m n/ if followed by /f v/ (nymph [nɪɱ̊ f], infant [ˈɪɱ̊ fənt], Tom Ford [ˈtɒɱ̊ ˈfɔːd]). b. Bilabial realisations [m] of /n/ before a bilabial consonant (in bed [ɪm ˈbed]). c. Velar realisations [ŋ] of /n/ before a velar consonant (ten coins [teŋ ˈkɔɪnz]). (4) Dental realisations [ ]̪ of alveolars /t d n l/ before a dental consonant /θ ð/ (tenth [ten̪θ], not those [ˈnɒt ̪ðəʊz], hide this [ˈhaɪd ðɪs], although [ɔl ̪ˈðəʊ]). For more details on the allophonic variants of RP sounds involving degree of openness and backness, in the case of vowels, and the place of articulation of consonants resulting from processes of coarticulation, please refer to the corresponding sections of Chapters 3 and 4 (Shockey 2003; Cruttenden 2014: 308–310; Gow et al. 2007).

5.2.8 Lenition and fortition The term fortition is used in phonology to refer to a strengthening in the overall force of a sound, whether diachronically or synchronically. Typically, fortition involves the change from a fricative to a stop, an approximant to a fricative, or a voiced to a voiceless sound. Lenition refers to the opposite phenomenon whereby the overall strength of a sound is weakened, as it occurs in the change from a stop to a fricative, a fricative to an approximant, a voiceless sound to a voiced sound, or a sound being reduced (lenite) to zero (Crystal 2008: 197, 274). English illustrations of both phenomena have already been introduced in Section 2.3.2.1 and are now summarised in the two following sections.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

242

Segment Dynamics: Aspects of Connected Speech

5.2.8.1 Affrication and fricativisation Plosives are said to be affricated when they are articulated with a slow, fricative release but in the same articulatory region. In this type of lenition, plosives have affricated variants that are followed by the brief corresponding fricative [pɸ b β ts dz kx gɣ], or they may even be replaced by the fricative sound in rapid speech, especially in intervocalic positions (imported [ɪmˈpɔˑtsɪd], [ɪmˈpɔˑsɪd]) (Shockey 2003; Collins and Mees 2009: 154; Cruttenden 2014: 172; Ashby 2011: 135; Buizza and Plug 2012). /p b/ are rarely affricated. Affrication is particularly associated with alveolar plosives, both in strongly accented positions (time [ˈtsaɪm], dime [ˈdzaɪm]), in weakly accented syllables /t d/ (to [tsə], reading [ˈriːdzɪŋ]), and in final positions (heart [hɑˑts]). Occasionally, with velar plosives affrication may also be heard, particularly in emphatic or hesitant speech in stressed phonetic contexts (come on [ˈkxʌm ɒn], good [ˈɡɣʊd]), or, more frequently, with /k/ in weakly accented or word-final syllables (black tea [blækx tsiː]. In English, the production of affricated plosives with homorganic fricative release regardless of the phonetic environment is a characteristic of regional pronunciation, especially of Liverpool (or Scouse) accent. A different type of lenition involves the fricativisation of approximants, as a result of which /r/ is realised as / [ɹ̝] (draft [dɹ̝ɑˑft] or [ɹ̥] (try [tɹ̥aɪ]), /l/ as [l ̥] (climb [kl ̥aɪm]), j as [j]̥ or [ç], and /w/ as [w̥ ], as already detailed in Section 4.2.5.

5.2.8.2 Glottalisation and glottaling Glottalisation is the cover term for any articulation involving a simultaneous glottal constriction, especially a glottal stop (plosive) (Crystal 1988: 213). The glottal stop [ʔ] is pronounced with the vocal folds coming into contact and with no vibration, so that there is a glottal closure and the lung air is compressed at that point. The airstream is released out of the oral cavity upon the sudden separation of the vocal folds. The compression stage consists of silence, which is auditorily perceived by the sudden cessation of the preceding sound or by the sudden initiation of the following sound. The preceding sound is shortened as it occurs with the other voiceless plosives. In RP, [ʔ] does not contrast with other plosives to produce differences in meaning, but rather it occurs as secondary articulation to fortis plosives and affricates. These are then said to be glottalised, or produced with a glottal reinforcement (or fortition), in which case a corresponding closure – and its corresponding release – is made either before or simultaneously with the oral closure of the plosive or affricate release (Christophersen 1955; O’Connor 1952; Roach 1973).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Co-articulation and allophonic variations of vowels and consonants

243

Pre-glottalisation85 is regularly used in RP to reinforce stressed syllableinitial vowels either to mark a syllable boundary or to add emphasis to the word, whatever the preceding stop, as shown in the examples below: geometry [dʒɪˈʔɒmətri] disorder [dɪsˈʔɔːdə] It is empty [ɪts ˈʔempti] I haven’t seen anybody [ˈaɪ ˈhævn̩t ˈsiːn ˈʔenibɒdi] An extended use of glottalisation in RP, though not subscribed to by conservative speakers, consists in reinforcing [p t k] when a vowel, nasal or lateral precedes them and they are followed by another consonant or a pause; affricates are also glottalised in theses contexts as well as between vowels: actor mat matches

[ˈæ͜ ʔktə] [mæ͜ ʔt] [ˈmæ͜ ʔʧɪz]

with glottalisation with glottalisation with glottalisation

[ˈæktə] [mæt] [ˈmæʧɪz]

without glottalisation without glottalisation without glottalisation

In the same phonetic contexts, glottaling may also occur. Regarded as the next stage to glottalisation, it involves the replacement of a fortis plosive, usually /t/ and less frequently /p k/, by a glottal stop, no oral closure being made (Fabricius 2002). The glottaling of /t/ in word-final positions and before syllabic [ɫ̩] and [n̩] is acceptable in Estuary English ( foot-rest [ˈfʊʔɹest], bottle [boʔɫ ̩], meet me [miːʔ miː]. Word-medially, however, the use of [ʔ] for /t/ remains stigmatised in RP (waiter [ˈweɪʔə], Britain [ˈbrɪʔn̩]). Glottaling of final /p k/ is not so common in RP. Its occurrence is restricted to homorganic consonant clusters within words and at close-knit word-boundaries (soap powder [ˈsəʊʔ ˈpaʊdə]). In Cockney and other regional (South East) urban dialects /t/ glottaling is heard in the same positions as RP, but more frequently and also in wider contexts such as word-medially and intervocalically (potato [pəˈteɪʔəʊ]). /p k/ may also undergo glottal replacements (supper [ˈsʌʔə], ticket [ˈtɪʔɪt]) in the same phonetic environments but they seem to retain their bilabial and velar closures. In addition, in Cockney speech [ʔ] may ocassionally replace fricatives such as final /f/ and initial /h/ (half a minute [ˈʔɑːʔ ə ˈmɪnɪʔ], You hate her [ju ʔeɪ ʔə]) (Gimson 1980: 168 ff; Finch and Ortiz Lira 1982: 56; Wells 1982: 341, 344, 347, 416; Roach 2005: 55; Cruttenden 2014: 182–186). Glottalisation and glottaling are essential features for SSLE (and other learners) to master. They are such prevalent features of RP that their correct use really makes a foreigner’s speech sound more native. 85 Tyneside is unique in showing post-glottalization, i.e. the addition of glottal reinforcement to a preceding plosive, in intervocalic positions (water [ˈwɔːtʔə]) (Cruttenden 2014: 185).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

244

Segment Dynamics: Aspects of Connected Speech

5.3 Assimilation and phonemic variations Assimilation of sounds is a process that takes place in connected speech whenever a sequence is particularly awkward to pronounce given its phonetic environment. We speak of assimilation when a phoneme is actually replaced by another phoneme as a result of phonetic conditioning usually at the wordboundary level, but also word-internally (Collins and Mees 2003: 203–216, 327; Collins and Mees 2009: 120–123, 274; Cruttenden 2014: 312–313). Assimilations occur both in formal and in informal speech, even though assimilated forms are more frequent in casual speech. Assimilations are blocked in slow speech and when the speaker makes a pause between words. There are different types of assimilation, attending to two parameters, which may interact in such a way that various possible combinations of assimilation processes are observed in connected speech, as will be shown in the following sections: (1) Direction of influence: one sound may affect either the preceding one or the following one; or two sounds may actually affect each other to the extent of both of them being replaced. The process is known as regressive assimilation (redressive, leading, or anticipatory) when sound 2 affects sound 1, or put differently, when a sound affects the preceding one so that the latter is anticipating articulatory features of the sound after it. The direction of influence is therefore backwards, from right to left, as in white plane [waɪp ˈpleɪn], where boldtype is used to highlight the spelling of the sound being assimilated. In contrast, the process is called progressive assimilation (lagging or perseverative) if sound 1 affects sound 2. In other words, a sound affects the following one so that the second sound carries articulatory features from the previous one. In this case the movement is from left to right, as in happen [ˈhæpm̩ ]. A third possibility is that two sounds may reciprocally affect each other to the extent of producing a different sound. This is a case of coalescent, or reciprocal assimilation, as in Would you? [wʊdʒu], giving place to the emergence of the affricate [dʒ]. (2) Type of influence: regardless of whether the assimilation is leading, lagging or reciprocal, this process may affect the assimilated sounds in different ways. Place of articulation is affected when the assimilated sound exhibits a different place of articulation from the original one. For example, an alveolar consonant may become bilabial in the vicinity of a bilabial sound. Manner of articulation can also be affected. The most common process is that of nasalisation, whereby a sound becomes nasal in the vicinity of another nasal sound. The quality of the voice can also be affected. This usually affects voiced sounds, which may become voiceless in the vicinity of voiceless sounds. We shall see that two of these changes may actually take place at the same time, such as place and manner.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Assimilation and phonemic variations

245

AI 5.3 Assimilation

5.3.1 Variations of place of articulation This process affects mostly alveolars, which become de-alveolarised and replaced by bilabials, velars or palate-alveolars, in a process which most native speakers are usually not aware of. Changes in the place of articulation can take place in leading, lagging or reciprocal directions, as we can see below.

5.3.1.1 Regressive assimilation Word-final alveolars /t, d, n, s, z/ adopt the place of articulation of the following consonants, without changing their voice, as can be seen in the following table, which summarises some of the most usual cases: Table 33: Regressive assimilation Articulation Sound 1

Followed by Sound 2

Regressive assimilation

New place of articulation

i. Alveolar /t/

Bilabial /p b m/

f

Bilabial [p]

/d/

/p b m/

f

[b]

add me [æb mi]

/n/

/p b m/

f

[m]

ten pens [tem penz] unplugged [ʌmˈplʌɡd]

ii. Alveolar /t/

Velar /k g/

f

Velar [k]

that car [ðæk kɑː]

/d/

/k g/

f

[g]

good guy [gʊg gaɪ]

/n/

/k g/

f

[ŋ]

ten cats [teŋ kæts]

iii. Alveolar /s/

Palato-alveolar /ʃ ʧ ʤ j/

f

Palato-alveolar [ʃ]

this yacht [ðiʃ jɒt]

/z/

/ʃ ʧ ʤ j/

f

[ʒ]

has Sheila? [hæʒ ˈʃiːlә]

Example

that pot [ðæp pɒt] white-board [ˈwaɪpbɔːd]

5.3.1.2 Progressive assimilation Changes in the place of articulation of phonemes in progressive processes are not very frequent, the most usual context being when a plosive is followed by a syllabic nasal, in which case the nasal may undergo assimilation to the place of articulation of the preceding plosive.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

246

Segment Dynamics: Aspects of Connected Speech

Table 34: Progressive assimilation Articulation Sound 1

Followed by Sound 2

Progressive assimilation

New place of articulation

Example

i. Bilabial /p b/

Nasal /n/

%

Bilabial [m]

cup and saucer [kʌp m ˈsɔːsә]

ii. Velar /k g/

Nasal /n/

%

Velar [ŋ]

bake and fry [beɪk ŋ fraɪ]

5.3.2 Variations in the manner of articulation. These are usually nasalisations, involving alveolars in very rapid speech, especially when close to the negative n’t. The following table summarises the most common regressive nasalisations: Table 35: Regressive nasalisations Articulation Sound 1

Followed by Sound 2

New manner of articulation

Example

/d/

/n/ /g/ /m/

[n] [ŋ] [m]

bad news [bæn njuːz] wooden gate [ˈwʊdŋ geɪt] good mum [gʊm mʌm]

/v/

/m/

[m]

of mine [әm maɪn]

5.3.3 Variations in voicing Changes in voicing are often due to regressive assimilations, whereby a wordfinal voiced fricative followed by a word-initial voiceless fricative may be realised as the corresponding voiceless fricative, as is illustrated in Table 36. Table 36: Regressive assimilations: voicing Articulation Sound 1

Followed by Sound 2

New manner of articulation

Example

/ð/ /z/ /v/

Voiceless Voiceless Voiceless

[θ] [s] [f]

with threats [wiθ θrets] it was supposed [ɪt wɒs sәˈpәʊzd] you’ve finished [juːf ˈfɪnɪʃt]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Assimilation and phonemic variations

247

Change of voicing is affected by progressive assimilation in the case of the /s/ and /z/ morphemes, represented by /S/ in Table 37 below, used in the formation of the plural of nouns, the genitive case, the third-person singular of the present tense of verbs, and also the /t/-/d/ ending of the past and participle of regular verbs, represented by /T/ below. In these cases, if the root of the word ends in a voiceless sound, then the morpheme assimilates to the voiceless variety, whereas a voiced consonant at the end of the root will have as an effect of an assimilation to the voiced variety as a realisation of the morpheme. Table 37: Progressive assimilations: voicing Archiphoneme and Morphemic meaning

Allomorphic or Morpho-phonemic variation

Example

{plural} /S/ {present} {genitive}

→ /s/ ‘voiceless’ / C [-voice] _ % → /z/ ‘voiced’ / C [+voice] _ %

cats, Pip’s, asks /kæts/ /pɪps/ /ɑːsks/ dogs, Bob’s, saves /dɒgz/ /bɒbz/ /seɪvz/

/T/ {simple past} {past participle}

→ /t/ ‘voiceless’ / C [-voice] _ % → /d/ ‘voiced’ / C [+voice] _ %

asked, parked /ɑːskt/ /pɑːkt/ saved, combed /seɪvd/ /kəʊmd/

5.3.4 Coalescent assimilation Coalescent assimilation is a cover term for the reciprocal assimilation of two underlying phonemes, usually involving place and manner assimilation at the same time, resulting in a third phoneme (Collins and Mees 2003: 328; Cruttenden 2014: 313). As shown in Table 38, in English coalescent assimilation is typically derived from the palatalisations of stops and fricatives, generally in close-knit word boundaries, as illustrated in Table 38 below. Table 38: Coalescent assimilation Articulation

Followed by Sound 2

New place and manner of articulation Sound 3

Sound 1

Example

i. Alveolar /t/

Palatal /j/

Palato-alveolar [tʃ ]

won’t you? [ˈwәʊnʧuː]

/d/

/j/

[dʒ]

did you? [ˈdɪʤuː]

/s/

/j/

[ʃ]

It’s your book! [ɪtʃɔː bʊk]

/z/

/j/

[ʒ]

Has your book come? [həʒɔː bʊk kʌm]

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

248

Segment Dynamics: Aspects of Connected Speech

One common case is when an unvoiced [ʧ] or voiced [dʒ] palato-alveolar affricate results from the fusion of correspondingly an alveolar unvoiced [t] or voiced [d] plosive and a palatal approximant [j], which becomes a voiceless [ ʃ ] or voiced [ʒ] palato-alveolar fricative, respectively (Harris 1994: 118). Hence, attention should be paid not to use [dʒ], for instance, when the sound preceding [j] is [t], as in let you, which in rapid speech should be pronounced as [ˈleʧu] not [ˈledʒu]. Another frequent example of coalescent assimilation also illustrated in Table 38 occurs when a voiceless /ʃ/ or voiced /ʒ/ palato-alveolar fricative results from the merging of [sj] and [zj], respectively.

5.4 Elision This phenomenon refers to the loss of sounds, particularly at word boundaries or very close to them (Collins and Mees 2009: 114; Cruttenden 2014: 313–314). Deletion of phonemes tends to occur both in formal and casual speech, and is more likely to happen the faster the speed of delivery is. Consonants are elided in the weak forms of some grammatical words: (1) Loss of /h/ in pronominal weak forms such as /ɪ/, /ɪm, /ә/ for he, him, her, or in some auxiliaries, such as have, has, had realised as /әv/, /әz/, /әd/, respectively. (2) Loss of /t/ in one of the weak forms of must /mәs/, as well as in other words in which syllable-final /t/ is preceded by another consonant of the same voicing and it is followed by a consonant other than [h] (last /lɑːst/, which may be pronounced as [lɑːs] with [t] elision). (3) Loss of /d/ in the weak form of and /әn/ and /n/; also in the weak form of does, which may be pronounced as /z/ or /s/. Syllable-final [d] elisions can also occur when this sound is preceded by another consonant of the same voicing and it is followed by a consonant other than [h], either across words or within a word (cold /kəʊld/ which may be pronounced as [kəʊl] with [d] elision). In addition, omission of alveolar plosives may also occur in the following contexts: (4) When word-final /t/ or /d/ are followed by initial either /t/ or /d/, the resulting cluster may be simplified by elision of the first plosive of the two, as in I’ve got to go /aɪv gɒtә gәʊ/, what does he like, /wɒdәz hi laɪk/, I could do it /aɪ kә du: ɪt/, he should talk /hi ʃә tɔːk/; (5) The /t/ of the negative /nt/ is often removed, especially in two-syllable words, when followed by a consonant, as in I shouldn’t carry it /aɪ ʃʊdn̩kæri ɪt/;

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Elision

249

(6) /t/ or /d/ may be elided when they occupy the middle position in a group of three consonants, most usually in the following contexts: (1) Voiceless continuant + /t/ or voiced continuant + /d/ (e.g. /st/, /ft/, /ʃt/, /nd/, /ld/, /zd/, /ðd/, /vd/) followed by a word-initial consonant (cf. finished late, hold tight, caused losses). (2) voiceless plosive or affricate + word final /t/ or voiced plosive + affricate + /d/ (/pt/, /kt/, /ʧt/, /bd/, /ɡd/, /ʤd/), followed by an initial consonant in the next word (e.g. picked one, looked fine, stopped speaking). This process may affect the past tense distinction, which is often resolved by contextual cues. Note that this elision process does not usually take place if the third consonant is /h/ or /j/, and that middle /k/, rather than /t/, is elided in the sequence /skt/, as in asked /ɑːst/. There are, however, some contexts in which plosives are not generally elided. Word-final /t/ and /d/ are retained before /h/ as in reached home /riːʧt hәʊm/ or round here /raʊnd hɪә/; or they may be kept in a coalesced form (/ʧ ʤ/), when followed by /j/ as in caught you /ˈkɔːʧә/ or lend you /ˈlenʤә/. Word-final /nt/ and /lt/ do not normally involve elision either. In this case, /t/ is normally replaced with /ʔ/ as in He spelt it wrong /hi spelʔ ɪt rɒŋ/. Other examples of consonant elision include: elision of /v/ in of before /ð/ (one of them /wʌn ә ðem/); elision of the consonant before /θ/ in numerals (sixth /sikθ/), and the elision of dental fricatives in some common nouns (clothes /klәʊz/, months /mʌnz/). Let us now have a look at vowel elisions. The most frequent instance is the elision of /ә/ occurring in two main contexts: (1) When it is preceded by a word-final consonant and followed by a continuant, which normally becomes syllabic, as in got another /gɒt ˈn̩әðә/; (2) When final /ә/ is followed by a linking /r/ plus word-initial vowel, as in after an hour /ˈɑːftrәn aʊә/. Elision may also affect the second element of a closing diphthong when followed by a syllable beginning with a vowel, whether word-internally (resulting in neutralisation, see § 5.7.2), as in tower /taʊә/ → /tɑːә/ → /tɑː/, or at word boundary, as in try another /traɪ әˈnʌðә/ → /tra әˈnʌðә/ → /trɑː ˈnʌðә/. It can be concluded, as noted by Ortiz and Lira (1982: 79), that the most frequent vowel elisions occur before /r/, /l/ or /n/.

AI 5.4 Elision

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

250

Segment Dynamics: Aspects of Connected Speech

5.5 Linking Linking (adjective / noun) or liaison (noun) are terms used in phonology to refer to one type of transition between sounds, where a sound is introduced at the end of a word if the following syllable has no onset, usually for ease of pronunciation (Crystal 2008: 280, 285). In English a word-final post-vocalic /r/ may be added to those words ending in /ɑː ɔː ɜː/ and /ә/, either single or in glides, when followed by a syllable beginning with a vowel, whether at morpheme or word boundary (Collins and Mees 2009: 118–120; Cruttenden 2014: 315–318) (see also Section 4.2.5.1). Now, this linking /r/ may be ‘historically justified’, i.e. it has existed in earlier stages of the language (e.g. her invoice), as indicated by the spelling, which is usually retained. This /r/ link is optional, although very frequent, between words, but pronounced when it occurs wordinternally, that is, the root of the word ends in /aː ɔː ɜː ә/ plus and the suffix begins with a vowel (e.g. occur + -ing > occurring /әˈkɜːrɪŋ/). Linking /r/ is present in both formal and informal speech. Linking /r/ should be distinguished from what is known as the “intrusive /r/”. This takes place when this linking function of /r/ is analogically extended to words in which the /r/ is not present in spelling at all (drawing /drɔːrɪŋ/), although the phonetic environment in which it occurs is similar to the one in which a linking /r/ arises. This link is very common after /ә/, as in vanilla ice /vәˈnɪlәr aɪs/. Intrusive /r/ is stigmatised as uneducated and therefore it should be avoided. In its place, some speakers may produce a glottal stop or a glide.

5.6 Juncture The phonetic boundary features between morphemes and words, which can help separate between otherwise identical units, are known as juncture (Cruttenden 2014: 318–319). The most common type of juncture is known as open or plus juncture, and it refers to that which takes place between words. In phonetic transcription open juncture is transcribed [+]. Let us examine one example to see how juncture features operate in order to distinguish between two otherwise phonemically identical sequences. Consider the sequences that stuff and that’s tough (taken from Crystal 2008: 258). A broad transcription illustrating connected speech will render both sequences as: /ðætstʌf/. However, if that stuff is meant, with the boundary between /t/ and /s/, the pronunciation of the initial /s/ will be powerful and the /t/ will be unaspirated; whereas if that’s tough is meant, the placement of the boundary

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Vowel gradation

251

between the /s/ and the /t/ means that the /s/, which is now final, will be less energetic, and the /t/, now in initial position, will be aspirated. Other well-known examples include: I scream-ice cream: /aɪ skriːm/ or /aɪs kriːm/; a name-an aim: /ə neɪm/ or /ən eɪm/; why chop-white shop: /waɪ tʃɒp/ or /waɪt ʃɒp/.

5.7 Vowel gradation The terms vowel gradation or ablaut are generally used in historical linguistics to refer to the relationship between verb forms based on variations in the root vowel, as in ring, rang, rung (Crystal 2008: 216). Here by “vowel gradation” we mean the process whereby some function words exhibit more than one quantitative and qualitative pronunciation pattern depending on whether they are unaccented (the most usual form) or accented (in special situations or said in isolation), and so they have correspond weak or strong pronunciations or forms depending on the phonetic context. In contrast, lexical words (nouns, adjectives, adverbs, main verbs) usually retain prominence based on the occurrence of a full vowel even when there is no pitch prominence, even if they are monosyllabic. This implies that they tend to have the same quantitative pattern in connected speech as in isolation.

5.7.1 Weak and strong forms In English weak pronunciations are much more frequent than strong ones, and so they should be regarded as basic, with occasional use of strengthening. However, it should be borne in mind that, while some function words have no weak form at all (on /ɒn/, off /ɒf/, when /wen/), others have more than one weak form. In the latter case the choice of one weak form over another usually depends on the phonetic context, but the effect of the phonetic environment may differ substantially from one accent or style to another. The 53 weak forms listed in Table 39 below, which can be compared with other lists proposed by Cruttenden (2014), Ortiz Lira (2008), García Lecumberri and Maidment (2000) or Kenworthy (1987) containing 55, 37, 44 and 35 entries, respectively, are typical of careful colloquial RP (Gimson 1980: 297) and are characterised by the following features: (1) Reduction in length (be [bi] instead of [biː]). (2) Vowel reduction or obscuration towards weak vowels (are [ә] instead of [ɑː]). (3) Elision of vowels (is [z] instead of [ɪz]).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

252

Segment Dynamics: Aspects of Connected Speech

(4) Elision of consonants (him [ɪm] instead of [hɪm]). Function words with elided /h/ are more common than the unelided ones (except in sentence-initial position, where only the latter are possible). For this reason SSLE are advised to pronounce and transcribe aitchless forms, as well as to counteract their tendency to mispronounce RP glottal fricative /h/ as a velar fricative [x]. (5) In those function words which have alternative weak pronunciations, one with a schwa/vowel and another one with schwa/vowel delition (as in the case of, for example, and, can, does, was and will), the former generally occurs before vowels and the latter before consonants. Table 39: Weak and strong forms WEAK basic

STRONG exceptional

a

[ә]

[eɪ]

am

[әm m]

[æm]

an

[әn]

and

[әnd әn

are

[ә]

[ɑː]

as

[әz]

[æz]87

at

[әt]

[æt]

be

[bi]

[biː]

been

[bin]

[biːn]

but

[bәt]

[bʌt]

[æn] n̩]86

can (aux. v) [kәn kn̩]88

[ænd]

[kæn]

could

[kәd]

[kʊd]

do89

[du dә] auxiliary verb

[duː] full verb / emphatic use of auxiliary verb

does

[dәz dz z s]90 auxiliary verb

[dʌz] full verb / emphatic use of auxiliary verb sometimes sentence initially

86 The pronunciation [ənd] is slightly more formal than [ən]. And is frequently reduced to [n̩] in rapid speech after [t d f v θ ð s z ʃ ʒ] ( fish and chips /ˌfɪʃ n̩ ˈtʃɪps/) 87 As generally has a weak pronunciation, the strong form is only used in sentence/utterance initial position when followed by an unstressed word. So compare As far as I am concerned /əz ˈfɑːr əz ˈaɪ əm kənˈsɜːnd/ with As I was tired /æz aɪ wz ˈtaɪəd/). 88 The weak realization of the modal auxiliary can is normally [kәn], but it may be pronounced [kn̩] if it is followed by a vowel (What can I do? /ˈwɒt kən aɪ ˈduː/ vs. What can you do? /ˈwɒt kn̩ ju ˈduː/). 89 The auxiliary verb do is pronounced [du] before vowels and [də] before consonants (Do I know you /du ˈaɪ ˈnəʊ ju/ vs. Do you know me? /də ju ˈnəʊ miː/). 90 Unless in emphatic contexts (Mary does know /ˈmeəri ˈdʌz ˈnəʊ/), the auxiliary form does is normally pronounced [dəz] before vowels (Does it sell well? /dəz ɪt ˈsel ˈwel/) and [dz] before consonants (What does he know /ˈwɒt dz hi ˈnəʊ/).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Vowel gradation

WEAK basic

253

STRONG exceptional

for

[fә]

[fɔː]

from

[frәm]

[frɒm]

had

[hәd әd d] auxiliary verb

[hæd] full verb / emphatic use of auxiliary verb

has

[hәz әz z] auxiliary verb

[hæz] full verb / emphatic use of auxiliary verb

have

[hәv әv v] auxiliary verb

[hæv] full verb / emphatic use of auxiliary verb

he

[hi iː ɪ]

[hiː]

her

[hә ә ɜː]91

[hɜː]

him

[ɪm]

[hɪm]

his

[ɪz]

[hɪz]

is

[s z]

[ɪz]

just

[dʒəst] when meaning ‘only’, ‘simply’ unstressed time adverb (with pres. perfect v)

[dʒʌst] when meaning ‘exactly’, ‘precisely’ stressed time adverb (with pres. perfect v)

me

[mi]

[miː]

must

[mәst mәs]92 when expressing ‘obligation’

[mʌst] when expressing ‘deduction’

not

[nt]

[nɒt]

of

[әv]

[ɒv]

our

[ɑː]

[ɑː aʊә]

or

[ә]

[ɔː]93

saint

[sәnt sәn sn] when followed by the name of a saint

[seɪnt] when the word occurs on its own

shall

[ ʃәl ʃl ̩ l ̩]94

[ ʃæl]

she

[ ʃi]

[ ʃiː]

should

[ ʃәd әd d]

[ ʃʊd]

91 These three weak forms have a linking-r when they are followed by a vowel. [ɜː] or [hә] are the usual pronunciations when her functions as a determiner (I brought her orchids /ˈaɪ ˈbrɔːt ɜː ˈɔːkɪdz/, while as an object it is usually pronounced [ə] or [hә] (I gave her the orchids /ˈaɪ ɡeɪv ə ði ˈɔːkɪdz/). 92 The weak form of must may be reduced to [mәs], with the [t] elided, when the next words begins with a consonant. 93 Or is generally pronounced [ɔː] and it has linking-r pronunciations when followed by a word beginning with a vowel (two or eleven /tuː ɔːr ɪˈlevn̩/). 94 Shall is often weakened to [ ʃәl] before a vowel (Shall I come? /ʃәl aɪ ˈkʌm) and to [ ʃl ̩] before a consonant (I shall come tomorrow ˈaɪ ʃl ̩ kʌm təˈmɒrəʊ).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

254

Segment Dynamics: Aspects of Connected Speech

WEAK basic

STRONG exceptional

sir

[sә] when followed by a name

[sɜː] when the word occurs on its own

some

[sәm sm̩ ] ‘an unspecified amount / number’

[sʌm] ‘a small / certain amount of something’ pronoun

than

[ðәn ðn̩]95

[ðæn]

that

[ðәt] conjunction and relative pronoun

[ðæt] demonstrative and adverb

the96

[ðə ði]

[ðiː]

their97

[ðə]

[ðeə]

them

[ðәm ðm̩ әm m]

[ðem]

there

[ðә] existential pronoun

[ðeә] adverb (homophone of their)

to

[tә tu]98

[tuː]

us

[әs s]

[ʌs]

was

[wәz]

[wɒz]

we

[wi]

[wiː]

were

[wә]

[wɜː]

who

[hu u] unstressed relative

[huː] NP (either stressed or not)

will

[wl ̩ l ̩]

[wɪl]

would

[wәd әd d]

[wʊd]

you

[ju]

[juː]

your

[jә] used only occasionally (LPD)

[jɔː] normal pronunciation

95 Than is generally pronounced [ðәn] before words beginning with a vowel (rather than eating /ˈrɑːðə ðən ˈiːtɪŋ/) and [ðn̩] before words starting with a consonant (shorter than legs /ˈʃɔːtə ðn̩ leɡz/). 96 The is pronounced [ðә] before words beginning with a consonant, and [ði] when it is followed by a vowel. The strong pronunciation [ðiː] is heard when the article is emphasised for some reason. 97 There normally takes the strong pronunciation [ðeə], but it can be weakened to [ðə], especially when followed by a word beginning with a vowel, and both forms have a linking-r when they are followed by a vowel (She caressed their arms /ʃi kəˈrest ðeər ɑːmz/ /ʃi kəˈrest ðər ɑːmz/). 98 The stranded version of to is usually [tu] rather than [tuː] (EPD). [tuː] is used restrictively in some fixed expressions such as to and fro, as well as in contrastive contexts (The letter was from him not to him /ðə ˈletə wəz ˈfrɒm ɪm nɒt ˈtuː ɪm/). In addition, it should be noted that to and into have two weak forms: [tə] and [ɪntə] are used before consonants (to/into the house /tə ˈɪntə ðə ˈhaʊs), while [tu] and [ɪntu] occur before vowels (to/into a house /tu ˈɪntə ə ˈhaʊs). In GA, however, these distinctions do not apply, because schwa pronunciations are heard in all contexts.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Vowel gradation

255

In rapid casual speech, there is a tendency for reduction and obscuration of vowels, both in grammatical words which do not usually have a weak form, and even in some lexical items as well. For example, the pronoun I in I didn’t know /ә/ or the possessive my in I love my baby /mә/). The words any and many, as in How many have you got? /haʊ mnɪ әv ju gɒt/. Concerning lexical words, these may show reduction of the short vowel (/ɪ ʊ ʌ æ ɒ/) or the diphthong /әʊ/ to /ә/, in unaccented positions close to a primary accent, as in You sit over there, with the verb sit as /sәt/ or He’s going to buy it, with going to as /ˈgәnә/. Strong forms, on the other hand, are produced in slow formal or solemn speech, as well as in the following contexts (Lillo 2009: 81–96): (1) When a function word is stressed for one or several of the following reasons: a. It is being quoted, as in Spell the word “was” /ˈspel ðә ˈwɜːd ˈwɒz/, where /ðә/ is weak, whereas /wɒz/ is used in its strong form. b. It is emphasised, as in I’d like milk and coffee, please /aɪd ˈlaɪk ˈmɪlk ˈænd ˈkɒfi |ˈpliːz/, where the speaker emphasises that s/he wants to have both milk and coffee. c. It is contrasted, as in The book isn’t for them, it’s for us /ðә ˈbʊk ˈɪzn̩t fә ˈðem | ɪts fәr ˈʌs/, where /ðә, ɪzn̩t, fәr/ are weak in this context, whereas /ðem/ and /ʌs/ exhibit their strong form for reasons of contrast. (2) When a preposition or auxiliary verb is stranded or is left on its own, with no object or full verb after it, usually in sentence-final position. This occurs in Where are you going to? /ˈweәr әr ju ˈgәʊɪŋ ˈtuː/, where /әr/ and /ju/ are weak in this context, whereas the preposition /tuː/ is strong as it comes in final position. Likewise, in the exchange A: Are you Mary? B: Yes, actually I am. /ə ju ˈmeəri | jes | ˈæktʃuəli ˈaɪ ˈæm/, the auxiliary /ˈæm/ has a strong pronunciation because it is stranded in utterance final position. However, stranded auxiliaries and prepositions may occur non-finally, but still in these cases they tend to have a strong pronunciation, as in I can, of course /ˈaɪ ˈkæn | əv ˈkɔːs/, or Peter was lauged at by his classmates /ˈpiːtə wəz ˈlɑːft ˈæt baɪ ɪz ˈklɑːsmeɪts/. (3) The negative contractions of (auxiliary and primary) verbs are always stressed and therefore have the same vowels as the strong form of their affirmative counterparts aren’t /ɑːnt/, couldn’t /ˈkʊdnt/, hadn’t /ˈhædnt/, hasn’t /ˈhæznt/, haven’t /ˈhævn̩t/, mustn’t /ˈmʌsnt/, shouldn’t /ˈʃʊdnt/, wasn’t /ˈwɒznt/, weren’t /wɜːnt/.

5.7.2 Neutralisation of weak forms As we have seen in section 5.8.1. and in Table 39, certain grammatical words may have different pronunciations depending on whether or not they are

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

256

Segment Dynamics: Aspects of Connected Speech

accented in the utterance (Cruttenden 2014: 305–307). This leads to confusion, since two words with different pronunciations, when said slowly or in isolation, may show the same pronunciation when said quickly in connected speech. This process of reduction to a common weak pronunciation is known as neutralisation. The consequent ambiguity is usually resolved by context and grammar. Tables 40 and 41 below (adapted from Gimson 1984: 279–80) show some of the most common neutralised forms and their full correspondences. Table 40: Neutralisation of weak forms Neutral form

Correspondences

Examples

/ә/

are, a

The bottles are new /ðә ˈbɒtlz ә njuː/ He bottles a new wine /hi ˈbɒtlz ә njuː waɪn/

a, her

She loves a girl /ʃi lʌvz ә gɜːl/ She loves her girl /ʃɪ lʌvz ә gɜːl/

or, of

Five or six people turned up /faɪv ә sɪks ˈpiːpl ̩ tɜːnd ʌp/ Five of them turned up /faɪv ә ðem tɜːnd ʌp

are, or

Hens are rare /henz ә reә bɜːdz/ Hens or rare birds /henz ә reә bɜːdz/

are, of

Three books of his /θriː bʊks ә hɪz/ Three books are his /θriː bʊks ә hɪz/

/әv/

have (aux.), of

A few have my books /ә fjuː әv maɪ bʊks/ A few of my books /ә fjuː әv maɪ bʊks/

/ðә/

there, the

There sounds as. . . /ðә saʊndz әz/ The sound is. . . /ðә saund әz/

/s/

is, has, does

It’s going /ɪts ˈgәʊɪŋ/ It’s gone /ɪts gɒn/ What’s (=does) it mean? /wɒts ɪt miːn/

/z/

is, has, does

Where’s she going? /weәz ʃiː ˈgәʊɪŋ/ Where’s she gone /weәz ʃiː gɒn/ Where’s she come from? /weәz ʃiː kʌm frɒm/

/әz/

as, has

As white as a sheet /әz waɪt әz ә ʃiːt/ Who has a sheet? /huː әz ә ʃiːt/

/әn/

and, an

Green and orange /grɪːn әn ˈorɪnʤ/ Eat an orange /iːt әn ˈɒrɪnʤ/

/n/

and, not

He hid and shut up /hi hɪd n̩ ʃʌt ʌp/ Didn’t he like it? /dɪd n̩ hiː laɪk ɪt/

/d/

had, would

She’d set my hair /ʃɪd set maɪ heә/ She’d set my hair /ʃɪd set maɪ heә/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

257

Advice to learners

Table 41: Neutralisation matrix a

an

and

A

as

does

had

has

have

is

not

of

or

the

there

/әn/

and

/әn/

/n/ /әr/

/ә/

as

/әz/

does

/s/ /z/

/s/ /z/

had

/d/

has

/әz/

/s/ /z/

/s/ /z/

have

/әv/

is not

/s/ /z/

/s/ /z/

/n/

of or

would

/ә/

an

are

are

/әv/ /әr/

the

/ðә/

there

/ðә/

would

/d/

5.8 Advice to learners Studies in second language acquisition of English have shown that, although certain connected speech phenomena such as velarisation (Martínez-Dauden and Llisterri 1990; Moore 2008) and palatalisation (Keys 2002), for example, may occur both in the L1 and the L2, it is not necessarily the case that they will be easily acquired in the second. Likewise, it has been observed that the application of the rules concerning connected speech phenomena, as in the case of secondary articulations, may be more difficult for some sounds or phonetic contexts than for others, much in the same way that connected speech phenomena can be more difficult to reproduce for some speakers than for others. At a more general level, it has thus been concluded that the acquisition of segment dynamics is partly affected by language-specific variation (Strange,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

258

Segment Dynamics: Aspects of Connected Speech

Weber, Levy, Shafiro and Nishi 2002), as well as by general production constraints generated by individual vocal tract configurations (Fowler and Saltzman 1993). Be that as it may, what previous work unanimously suggests is that nonnative speakers must learn the general parameters of connected speech variability from the language environment. Foreign learners often learn English pronunciation on a basis of isolated word forms. However, in conversations and specially in colloquial speech, native speakers frequently modify words as described in the foregoing sections. It may not be necessary for SSLE to reproduce all the special context forms that have been described, but those aiming at a native speaker competence should be aware of the features that characterise more casual pronunciation, in particular the proper use of weak forms, liaisons, elisions and the co-articulatory and assimilatory tendencies (especially those of [n], [t], [d], [s] and [z]) that apply in certain contexts given the high frequency with which these processes occur in English. Nevertheless, elisions should be avoided when the dropping of a sound triggers the loss of a grammatical or semantic meaning (tense, number, possession, etc.). An excess of pre-vocalic glottal stops should also be discouraged (Cruttenden 2014: 321; Estebas Vilaplana 2009: 118–127).

Further reading Readers wishing more detail on connected speech phenomena and other related issues such as transcription and segmentation should consult, for example, Cruttenden (2014, Chapters 11 and 12), Ashby (2011, Chapter 9), (Lodge 2009, Chapters 4 to 7), Ogden (2009, Chapter 3), and Ladefoged (2005, Chapter 16). In addition, we also recommend Shockey (2003) because it focuses on the phonetic and phonological features that characterise conversational English such as reduction, syllabicity and weak forms, among others, and it also presents experimental studies that illustrate how casual speech is produced and perceived. Turning to long established texts particularly addressed to SSLE, descriptions of connected speech phenomena may be found in Monroy Casas (1980, 1981, 2012, Chapter 2), Estebas Vilaplana (2009, Chapters 4), Mott (2011, Chapter 6), Finch and Ortiz Lira (1982, Chapter 8), Sánchez Benedicto (1980) and Stockwell and Bowen (1965), among others.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Advice to learners

259

Exercises 1.

2.

Render the following text given in narrow (phonetic) transcription into ordinary spelling.99 [ð̥eːːɪ ɔːɫ ̥ l ʲ̥ ɪvd̥ tʰəˈɡ̟eðəɹ̥ ɪ ̃n̥ ə sm̠ ʷɔːɫ ̥ haˑʊs ɪ ̃n̥ ð̥ə ˈv̥eɹʲi ˈsẽnt̬ʰəɹ̥ ɒv̥ ə d̥ɑˑk pʰaːɪ ̃n̥ w̥ ʊd̥ // ˈɪ ̃nt̬ʰə ð̥ ʲ ɪs pˡleˑɪs ð̥ə sʌ̃n̥ ˈn̥ evə ʃɒ̃n̥ b̥ ʲ ɪˈk̬əz̥ ð̥ə ʃeːɪd̥ʷ w̥ əz̥ t̠ʰʷuːː d̥ ʲ iˑp / ə̃n̥ n̥ əʊ w̥ ʲɪ ̃nd̥ ˈevə k̟ʰeːɪ ̃m̥ ð̥eə ˈaːɪðə / b̥ ʲ ɪˈk̬əz̥ ð̥ə b̥aːʊz̥ w̥ ə t̠ʰʷuːː θʲɪk / səːːʊ ð̥æt ɪt w̥ əz̥ ð̥ə m̥ əˑʊst ˈs̠ʷɒlʲɪt̬ɹ ʲ̩ i ə̃n̥ kʰʷwaˑɪət pˡleˑɪs ɪ ̃n̥ ð̥ə w̥ ɜːld̥ / ə̃n̥ ð̥ə fʲɪˈl ̠ʷɒsə̬ f̬əz̥ w̥ əɹ̥ ˈeːɪbˡɫ ̩ tʰə h̟ ʲ ɪəɹ̥ iˑʧ ˈʌðə ˈθʲɪ ̃ŋk̟ʰ ʲɪ ̃ŋ̊ ɔːɫ ̥ d̥eːːɪ lʷɒ̃ ̥̠ ŋ̊ / ɔːː ˈm̥ eˑɪk̟ʰ ʲɪ ̃ŋ̊ ˈspʲiˑʧ̬ ʲɪz̥ tʰə iˑʧ ˈʌðə //] Allophonic variation of vowels. Classify the following words under the right heading: normal length, clipping, lengthening and extra length. code wool pea tear breath

3.

steep coat search through sang

pit soothe claim big enough

bed turn cord wound blob

higher thief niece south car

Allophonic variation. Give the allophonic transcription of the following words: appeal Spain care of

vocalic acknowledge behave

accomplish leaked it took

middle thank you cold

pointing acclaims either of

4. Render the following text into narrow (phonetic) transcription, paying special attention to all the allophonic variations of both vowels and consonants. When the little boy found that the light did not move he drew closer to it, and at last, emboldened by curiosity, he stepped right into it and found that it was not a thing at all. The instant that he stepped into the light he found it was hot, and this so frightened him that he jumped out of it again and ran behind a tree. Then he jumped into it for a moment and out of it again, and for nearly half an hour he played a splendid game of tip and tig with the sunlight. At last he grew quite bold and stood in it and found that it did not burn him at all, but he did not like to remain in it, fearing that he might be cooked.100

99 The Project Gutenberg EBook of The Crock of Gold, by James Stephens http://www. gutenberg.org/files/1605/1605-h/1605-h.htm. 100 Same as in Exercise 1.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

260 5.

Segment Dynamics: Aspects of Connected Speech

Assimilation. Identify instances of assimilation in the following sequences and explain each case. Try and pronounce them without assimilation. Does it seem to affect speech speed? 1. white paper 5. in question 9. newspaper

6.

2. stand by 6. seven pots 10. good cook

3. bath salts 7. that ghost

4. gold plate 8. nice shells

Elision. Identify possible cases of elision in the following words and sequences. Explain them. Now try and pronounce the words without the elision: does it sound natural? 1. stand there 5. stand here 9. picked two

2. I don’t know 6. stand aside 10. give me one

3. next please 7. picked one 11. but came

4. postman 8. asked him 12. can’t hurry

7.

In the following passage given in orthography, try and identify the places where either linking or intrusive r might occur and comment on each case. In these people the children were deeply interested. They used to go apart afterwards and talk about them, and would try to remember what they looked like, how they talked, and their manner of walking or taking snuff. After a time they became interested in the problems which these people submitted to their parents and the replies or instructions wherewith the latter relieved them. Long training had made the children able to sit perfectly quiet, so that when the talk came to the interesting part they were entirely forgotten, and ideas which might otherwise have been spared their youth became the commonplaces of their conversation.101 8. In the following passage, given in orthography, try and identify all the grammatical words susceptible of use in their weak form. Comment on whether the weak or the strong form is used and why. Brother, these are weighty reflections, and I do clearly perceive that the time has come for you to stop. I might observe, not in order to combat your views, but merely to continue an interesting conversation, that there are still some knowledges which you have not assimilated – you do not yet know how to play the tambourine, nor how to be nice to your wife, nor how to get up first in the morning and cook the breakfast. Have you learned how to smoke strong tobacco as I do? or can you dance in the moonlight with a woman of the Shee? To understand the theory which underlies all things is

101 Same as above.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Advice to learners

261

not sufficient. It has occurred to me, brother, that wisdom may not be the end of everything. Goodness and kindliness are, perhaps, beyond wisdom. Is it not possible that the ultimate end is gaiety and music and a dance of joy? Wisdom is the oldest of all things. Wisdom is all head and no heart. Behold, brother, you are being crushed under the weight of your head. You are dying of old age while you are yet a child.102

102 Same as above.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:11 PM

Chapter 6

6 Beyond the Segment: Stress, Rhythm and Intonation 6.1 Introduction So far, we have dealt with what is usually referred to as segmental phonetics and phonology, that is, the articulatory, acoustic and auditory properties of discrete segments or speech sounds and the phonological processes involved in their production, paying particular attention to the sounds of RP in comparison with those of PSp. It is time now to go beyond the individual segment in order to study the features that extend over and above a sequence of segments, occurring at a higher level of the utterance. These features are known as supra-segmental or prosodic, namely, stress at sentence level, rhythm and intonation. Prosodic units are chunks of speech, which may not coincide with grammatical units, whose boundaries are marked by the breathing processes of inhalation and exhalation, gradually delineating a contour in which pitch declines and vowels lengthen over the duration of the unit, until both pitch and speed are reset at the boundary of the next unit. This is what gives languages and dialects their idiosyncratic prosody or melody. In this chapter we will explore their phonetic realisation and perception and how they fit into the phonological system of English – i.e. how they can affect meaning (Collins and Mees 2003: 226–275; Collins and Mees 2009: 123–148; Roach 2005: 93–203; Crystal 2008: 393; Cruttenden 2014: 239–269).

6.2 Stress and rhythm Stress is a relative concept, which may be defined in terms of production and reception, as the degree of force, or relative emphasis, which is given to a syllable in a word in isolation (word stress or lexical stress) or to certain words within a longer utterance (prosodic stress or sentence stress),103 and which means that stressed syllables are then perceived as more forceful than others. However, it must be noted that prominence does not simply imply a more forceful articulation, i.e. an increase in loudness, but it is due to a concurrence of other phonetic features, such as pitch, length of the vowel in the stressed syllable, and subsequent changes in the quality of not only the vowels affected, but also the consonants and their articulation. In this unit, we are going to see that when 103 We avoid the use of the term ‘sentence’ to refer to the stress words take in connected speech, so as not to link the notion of ‘utterance’ in general to a particular grammatical category.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

263

words are put together in utterances, they are subject to rhythmical variations, usage conventions and the intentions of the speaker. This means that words which in isolation would always be stressed, as seen in Chapter 5, may lose their stress when they are considered in connected speech because of the presence of stronger stresses in surrounding syllables; and likewise, words which would normally not be stressed, may be stressed for a certain effect (Finch and Ortiz Lira 1982: 87–90; Crystal 2008: 454–456; Roach 2005: 93–103). 6.2.1 Stress-timed languages versus syllable-timed languages Stress and rhythm are closely connected. Rhythm is the periodic repetition of an event, that is, at regular intervals of time. English is a stress-timed language whereas Spanish is syllable-timed (Crystal 2008: 456; Collins and Mees 2003: 241–242; Cruttenden 2014: 271–272). The difference between these two rhythms may be comparable to what in music is known as legato, or ‘tied together’, and staccato or ‘detached’. This means that in English stressed syllables recur at equal intervals of time, regardless of the number of unstressed syllables between them, which normally get shortened or phonetically weakened as they are fitted in quick succession between stresses. This can be seen in two different ways by considering the following examples: (1)

ˈSusan had ˈbought her the ˈapples in the ˈmarket. || ˈsuːzn̩ həd ˈbɔːt hə ði ˈæpl ̩z ɪn ðə ˈmɑːkɪt ||

(2)

a. ˈJohn ˈcame ˈhome at ˈfive. || ˈdʒɒn ˈkeɪm ˈhəʊm ət ˈfaɪv || b. The ˈboys aˈrrived at the ˈschool at ˈseven. || ðə ˈbɔɪz əˈraɪvd ət ðə ˈskuːl ət ˈsevn̩ ||

In example (1), we can observe that since the stresses occur at equal intervals, the syllables in between have to take roughly the same time to be pronounced, even though there are only two syllables between su- and bought or bought and a, and yet there are three syllables between a and mar and only one after mar. This is what is meant by stress-timed: stress rules the time taken to pronounce unstressed syllables. In examples (2a) and (2b) we see that although (2a) contains a total of five syllables and (2b) has exactly double, the two utterances should take roughly the same amount of time to pronounce, since they each contain four stresses. A side-effect of English stress-timed rhythm is vowel weakening, whereby vowels are shifted to the central part of the CVS (Carr 2012: 33) in unstressed syllables resulting in the production of weak vowels /ɪ ʊ e ə/ that are located in that area. Thus, it can be seen in the phonemic transcriptions of (1) and

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

264

Beyond the Segment: Stress, Rhythm and Intonation

(2) above, as well as in all previous and subsequent ones, that most unstressed syllables have one of these weak vowels, mostly /ə/. Syllable-timing, on the other hand, means that syllables are repeated periodically, taking approximately the same time, regardless of whether they are stressed or not, producing the “machine gun” effect of Spanish speech. To give similar examples to the English ones above, consider (3) and (4) below: (3) Suˈsana venˈdrá el ˈsábado a las ˈcinco. ‘Susan will come on Saturday at five.’ (4) a. ˈElla ˈcanta ˈbien. ‘She sings well.’ b. Los ˈtigres y aˈquellos eleˈfantes ‘The tigers and those elephants’ In (3) because there is one syllable between drá and sá and there are four syllables between sá and cin, the latter sequence will take more time to pronounce than the former. For this same reason, (4a) with five syllables will take less time to pronounce than (4b), with ten, even though they both exhibit the same number of stresses: in this case, the rhythm is syllable-driven and, as a result, no vowel weakening takes place. Therefore, it is extremely important for Spanish speakers to pronounce English utterances paying attention to phonic groups, trying to join together unstressed syllables with stressed ones into a single one, in a legato manner. 6.2.2 Word stress We have mentioned the phonetic features of stress, but now we must consider whether stress has a phonemic function serving to distinguish words that otherwise would be homophones. Lexical stress is phonemic in English and Spanish. For example, the noun ˈincrease and the verb inˈcrease are distinguished by the positioning of the stress on the first syllable in the former, and on the second syllable in the latter. Likewise, many pairs of compound words and word sequences are distinguished by means of stress (blackbird vs black ˈbird; ˈpaper bag vs paper ˈbag); and to provide further examples of English and Spanish, words such as insight and incite in English, or in the Spanish compro and compró are only distinguished by the position of the stress – that is, the stress falling on the first syllable in the former and on the second in the latter, respectively. We are going to look at stress placement within the word in isolation, what is known as citation form, which is not the way we normally speak words aloud. By doing this, we can focus more clearly on the different levels of stress

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

265

within the word, before we see how this may be affected when words are spoken together in connected speech. 6.2.2.1 Levels of stress It is possible to distinguish up to five different levels of stress within a polysyllabic word, but for practical purposes we are going to distinguish only three, which is what most dictionaries do: primary, secondary and unstressed. Every polysyllabic word has a primary stressed syllable and everything else remains unstressed, with another syllable possibly carrying secondary stress, which is not always predictable. The standard notation of stress in transcription, a superscript straight comma before the accented syllable with a primary stress, and a subscript one for secondary stress as is already familiar in this book. In monosyllables the stress is not generally marked. (Collins and Mees 2003: 229–230; Roach 2005: 95–100). 6.2.2.2 Placement of stress The position of the stress in a given word is not as predictable in English or Spanish as it is in other languages, like Czech or Polish, for example, which have clearer guidelines or rules. However, certain rules seem to be in operation given the fact that native speakers have no difficulty in assigning the right stress to an unfamiliar word. Such guidelines work on the basis of four parameters, which can help identify where to place the stress in any given word in English. These are: (1) the morphological structure of the word: is it simple, i.e. containing only one grammatical unit, or complex, either by derivation with prefixes and/or suffixes or compounding? (2) the grammatical category of the word (noun, adjective, verb, etc.) (3) the number of syllables of a word. (4) the phonological structure of those syllables. However, any attempt at a complete prescriptive system of all the rules deriving from the application of the above parameters would be much too complex in the context of foreign language learning. For that reason, while still following the above considerations, we will be looking at the most frequent patterns which may prove useful for students. A first division of words focuses on word length, two and three syllable words being considered short on the one hand, and words with more than three syllables being considered long, regardless of their morphological structure. We want to avoid morphological considerations as much as possible, which in turn have a lot to do with etymology and the origin of words, as they would complicate the general picture unnecessarily for present purposes. These rules apply to major lexical words only.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

266

Beyond the Segment: Stress, Rhythm and Intonation

AI 6.1 Stress in polysyllables

(1) Short words (maximum three syllables) As a very general rule, there is a general tendency for stress to fall on the first syllable of nouns with two or three syllables; however, if that syllable is weak then the stress is shifted to the next syllable on the right. The tendency for simple verbs and adjectives of two or three syllables is different as they tend to carry stress on the final syllable, unless this is weak, in which case stress will be placed on the penultimate syllable (second from the end); if this is also weak, stress will then be shifted to the first syllable. Note that the diphthong /әʊ/ is never stressed when final (Finch and Ortiz Lira 1982: 90–94; Collins and Mees 2003: 231; Roach 2005: 96–100). Table 42: General tendencies in stress patterns: 2 and 3 syllable words 2 syllables

3 syllables

Nouns

ˈbottle ˈmorning

ˈprinciple ˈquality

Verbs

aˈpply, aˈttract,

conˈsider conˈtribute

Adjectives

aˈlone unˈknown

imˈportant

Following this general tendency, quite a few words in English which can function as both verbs and nouns exhibit this contrastive stress pattern: nouns tend to be stressed on the first syllable and verbs on the last. This is known as variable stress or stress switch. Note that in this case those words functioning as adjectives are stressed on the first syllable as well, not on the second, to distinguish them from verbs. This stress switch may also be associated to a change in the quality of the vowels (Cruttenden 2001: 233; Roach 2009: 87).104

AI 6.2. Variable stress (2) Long words (more than three syllables) A very strong tendency is for long words in general to carry the stress on the antepenultimate syllable – i.e. the third from the end- as in e ˈmergency, an 104 Some exceptions are: ˈcomment (vb), a ˈddress (n.), reˈsearch (n.). Note also: ˈminute (n.) but mi ˈnute (adj.), ˈinvalid (n) but inˈvalid (adj.).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

Table 43: Most common pairs of words with variable stress

accent abstract attribute compact compound concert conduct conflict construct content contest contract contrast convert convict decrease desert dictate escort exploit export extract frequent import incense increase insert insult invite object perfect perfume permit pervert produce progress project protest rebel record refund refuse reject segment subject torment transfer transplant transport update

Verbs

Nouns/Adjectives

/ækˈsent/ /æbsˈtrækt/ /æˈtrɪbjuːt/ /kəmˈpækt/ /kəmˈpaʊnd/ /kənˈsɜːt/ /kənˈdʌkt/ /kənˈflɪkt/ /kənˈstrʌkt/ /kənˈtent/ /kənˈtest/ /kənˈtrækt/ /kənˈtrɑːst/ /kənˈvɜːt/ /kənˈvɪkt/ /dɪˈkriːs/ /dɪˈzɜːt/ /dɪkˈteɪt/ /ɪˈskɔːt/ /ɪkˈsplɔɪt/ /ɪkˈspɔːt/ /ɪkˈstrækt/ /frɪˈkwent/ /ɪmˈpɔːt/ /ɪnˈsens/ /ɪnˈkriːs/ /ɪnˈsɜːt/ /ɪnˈsʌlt/ /ɪnˈvaɪt/ /əbˈdʒekt/ /pəˈfekt/ /pəˈfjuːm/ /pəˈmɪt/ /pəˈvɜːt/ /prəˈdjuːs/ /prəˈɡres/ /prəˈdʒekt/ /prəˈtest/ /rɪˈbel/ /rɪˈkɔːd/ /rɪˈfʌnd/ /rɪˈfjuːz/ /rɪˈdʒekt/ /seɡˈment/ /sʌbˈdʒekt/ /tɔːˈment/ /trænsˈfɜː/ /trænsˈplɑːnt/ /trænsˈpɔːt/ /ʌpˈdeɪt/

/ˈæksent/ /ˈæbstrækt/ /ˈætrɪbjuːt/ /ˈkɒmpækt/ /ˈkɒmpaʊnd/ /ˈkɒnsət/ /ˈkɒndʌkt/ /ˈkɒnflɪkt/ /ˈkɒnstrʌkt/ /ˈkɒntent/ /ˈkɒntest/ /ˈkɒntrækt/ /ˈkɒntrɑːst/ /ˈkɒnvət/ /ˈkɒnvɪkt/ /ˈdiːkriːs/ /ˈdezət/ /ˈdɪkteɪt/ /ˈeskɔːt/ /ˈeksplɔɪt/ /ˈekspɔːt/ /ˈekstrækt/ /ˈfriːkwent/ /ˈɪmpɔːt/ /ˈɪnsəns/ /ˈɪnkriːs/ /ˈɪnsət/ /ˈɪnsʌlt/ /ˈɪnvaɪt/ /ˈɒbdʒɪkt/ /ˈpɜːfɪkt/ /ˈpɜːfjuːm/ /ˈpɜːmɪt/ /ˈpɜːvɜːt/ /ˈprɒdjuːs/ /ˈprəʊɡres/ /ˈprɒdʒekt/ /ˈprəʊtest/ /ˈrebəl/ /ˈrekɔːd/ /ˈriːfʌnd/ /ˈrefjuːs/ /ˈriːdʒəkt/ /ˈseɡmənt/ /ˈsʌbdʒɪkt/ /ˈtɔːment/ /ˈtrænsfə/ /ˈtrænsplɑːnt/ /ˈtrænspɔːt/ /ˈʌpdeɪt/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

267

268

Beyond the Segment: Stress, Rhythm and Intonation

ˈticipate, iˈronical. In words with three or more syllables secondary stress is frequent, usually when the primary stress falls on the third syllable or later. Here it is normal for secondary stress to be marked on one of the preceding syllables (normally not the one immediately preceding the primary stress). For example, ˌinter ˈjection and ˌevo ˈlution, which have their primary stress on the third syllable, consequently take secondary stress on their first syllables; ˌorgani ˈzation and a ˈssoci ˈation both have primary stress on the fourth syllable, but have secondary stress on the first and second syllables respectively. If the primary stress falls on the third or fourth syllable from the end, that is, quite early in the word, the secondary stress will then fall on a later syllable: ˈMedite ˌrranean. Although so far we have been considering short and long words in general, it may prove useful to pay some attention to their morphological structure, since some very general guidelines for stress placement can be observed. Very briefly, a word may become complex by means of two different processes: by affixation of either a prefix before the basic form of the word, or stem, as in in-adequacy, or a suffix after it, as in contrast-ive; and also by compounding, of two independent stems, as in basket-ball. In the case of prefixes, a very general tendency is for primary stress to fall on the syllable following the prefix, as in de ˈcrease ex ˈclaim, i ˈllegal, re ˈform and un ˈreliable, for example -except for a large number of nouns and adjectives, such as ˈabstract, ˈcontract, ˈexchange or ˈprodigal. Suffixation is slightly more complex, but since most suffixes are easily recognisable by students, it seems worthwhile to have a look at their most frequent stress patterns. Suffixes may be divided in two groups: (a) those that do not affect the stress pattern of the stem, in (3), and (b) those that cause stress shift of some sort, exemplified in (4). Note that for practical purposes single and double endings are not differentiated, and that stems may not necessarily be obvious “English” words, but vestiges of older words of other origins, like Latin or Greek. Note also that there will be a secondary stress on a previous syllable in the case of long words (Finch and Ortiz Lira 1982: 94–97; Collins and Mees 2003: 232–237; Roach 2005: 104–111; Cruttenden 2014: 242–248).

AI 6.3 Suffixes which do not affect the pronunciation of the stem (3) Suffixes that do not affect the stress pattern of the stem

ˈlaughable, ˈperishable, reˈmarkable, unforˈgettable

ˈcarriage, ˈmarriage, ˈpackage, perˈcentage

ˈfloral, ˈordinal, ˈnatural, ˈnaval

ˈjuicy, ˈcurrency, ˈdelicacy, ˈintimacy

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm















Note:

269

archˈbishopdom, ˈfreedom, ˈkingdom, ˈprincedom ˈchristen, enˈlighten, ˈmadden, ˈwiden ˈgardener, ˈlecturer, ˈplumber, ˈsilencer ˈbeautiful, ˈharmful, ˈgraceful, forˈgetful ˈclassify, ˈglorify, iˈdentify, ˈnotify ˈchildhood, ˈfalsehood, ˈlikelihood, ˈwomanhood ˈboring, conˈfirmingˌ enterˈtaining, imˈproving aˈpologise, ˈcivilize, ˈnormalize, rˈealise exˈtremist ˈfinalist ˈjournalist ˈoculist ˈchildlike ˈfactorylike ˈlifelike ˈwarlike ˈaimless, ˈbottomless, comˈpassionless, ˈuseless ˈbooklet, ˈeyelet, ˈpiglet, ˈquadruplet ˈduckling, ˈlordling, ˈyearling biˈlingually, iˈdeally, ˈlovely, ˈscholarly aˈgreement, ˌdisaˈppointment, ˈpunishment, ˈvestment ˈcalmness, ˈcheerfulness, ˈfaithfulness, ˈsadness auˈtonomous, conˈtemptuous, ˈdangerous, ˈvigorous ˈcitizenship, ˈfellowship, ˈfriendship, ˈscholarship ˈcapture, deˈparture, ˈfailure, ˈsignature ˈclockwise, ˈlengthwise, ˈlikewise, ˈotherwise ˈairy, ˈrocky, ˈfunny, ˈnoisy

(4) Suffixes which affect stress placement: two groups a. Suffixes which tend to attract the primary stress to the syllable preceding them; this means that the original primary stress may be shifted to the last syllable of the stem, just before the suffix. There may also be changes in the quality of vowels.

AI 6.4 Stress shifting suffixes





aˈffirmative, comˈparative, deˈmonstrative, imˈperative deˈficient, eˈfficient, proˈficient, suˈfficient coefˈficiency, deˈficiency, improˈficiency, suˈfficiency ˌadvanˈtageous, couˈrageous, misceˈllaneous, ˈrighteous ˌbibliˈography, ˌcinemaˈtography, disˈcography, phoˈtography adˈverbial, coˈllegial, confiˈdential, meˈmorial Augusˈtinian, Caˈnadian, cosmeˈtician, dictaˈtorian alcoˈholic, Aristoˈtelic, cateˈgoric, magˈnetic

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

270

Beyond the Segment: Stress, Rhythm and Intonation











acaˈdemical, aposˈtolical, matheˈmatical, synˈtactical ˈcriminal, meˈdicinal, oˈriginal, ˈvirginal adminisˈtration, compliˈcation, conˈception, exˈtension laˈborious, inˈfectious, luˈxurious, reˈligious ˈcapital, ˈorbital, reˈcital, ˈvital caˈlamitous, feˈlicitous, forˈtuitous, uˈbiquitous ˈaltitude, f ˈortitude, ineˈxactitude, ˈmultitude deˈfinitive, ˈprimitive, reˈpetitive, ˈsensitive aˈcidity ambiˈguity Chrisˈtianity seˈxuality conˈtemplative, diˈgestive, imˈpressive, reˈflexive biˈlingual, conˈtextual, eˈventual, spiˈritual moˈlecular, parˈticular, quaˈdrangular, ˈcircular conˈtinuous, imˈpetuous, inconsˈpicuous, temˈpestuous ˈbackwards ˈforwards ˈhomewards ˈoutwards

b. Suffixes which attract the primary stress to themselves, that is, the original primary stress is shifted to the actual suffix.

AI 6.5 Stress carrying suffixes













cruˈsade, lemoˈnade, masqueˈrade, oranˈgeade doˈmain, orˈdain, planˈtain, teˈrrain diploˈmatic, dogˈmatic, idioˈmatic, systeˈmatic accuˈsation, circuˈlation, derivˈation, falsifiˈcation absenˈtee, briˈbee, interˈviewee, refuˈgee auctioˈneer, domiˈneer, mountaiˈneer, puppeˈteer fundaˈmental, governˈmental, instruˈmental, occiˈdental eviˈdential, influˈential, resiˈdential, toˈrrential efferˈvesce, fluoˈresce, lumiˈnesce, phosphoˈresce Bologˈnese, Cantoˈnese, Chiˈnese, Japaˈnese araˈbesque, burˈlesque, pictuˈresque, romaˈnesque disˈquette, kitcheˈnette, luˈnette, vinegaˈrette apoloˈgetic, geˈnetic, homiˈletic, paˈthetic anˈxiety, proˈpriety, soˈciety, vaˈriety anˈtique, criˈtique, mysˈtique, uˈnique acquiˈsition, compoˈsition, eˈdition, prepoˈsition baˈlloon, ponˈtoon, raˈtoon, saˈloon herˈself, itˈself, ourˈselves, themˈselves absoˈlution, constiˈtution, dissoˈlution, soˈlution

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

271

(5) Compounds Let us now focus on compounding. Knowing the general tendencies which operate in the stressing of compounds will make it much easier for the student to sound more native-like. A compound word is one which is made up of two (less frequently three) independent words, which may be written as one word or as two, with or without a hyphen, and which act together as one grammatical and semantic unit: blackbird, week-end, paper clip. The grammatical structure of the word, together with certain lexical aspects, will prove very useful in guessing whether the stress falls on the first or the second element. Compounds with the stress on the first element tend to be single-accented, like those presented in (5.1), whereas compounds with the stress on the second element tend to have a secondary stress on the first, which are listed under (5.2). An important prosodic phenomenon that affects compounds and long words in English is stress shift. We have mentioned above that English is a stresstimed language, unlike Spanish, which is syllable-timed. Because of this, two stressed syllables in a row are avoided, increasing the distance between stresses whenever possible. This means that, as we shall see later when dealing with the nuclear stress of an utterance, compounds, and other long words, with a primary stress on the last syllable and an earlier secondary stress will shift the primary stress to the first syllable in the vicinity of another stress in the utterance. For example: consider the compound New York in: (5) I am in New ˈYork and

ˈNew York ˈCity is huge

(5.1) Compounds that have the primary stress on the first element: a. Most of those made up of two single-syllable words, especially when written as one word: ˈairport, ˈfootball, ˈblackboard, ˈcowboy. b. Those in which the second element indicates the performer of the action, of the pattern “noun+agent noun”: ˈtypewriter, ˈbaby-sitter, ˈpainkiller, ˈrecordplayer. c. Those in which the first noun delimits the meaning of the second, indicating its purpose or “what type of thing”, of the pattern “–ing verb/noun + noun”: ˈchewing ˌgum, ˈhearing ˌaid, ˈfiling ˌcabinet, ˈnail ˌpolish, ˈtimeˌtable, ˈnews ˌpaper, ˈtoothˌbrush. d. Those nouns with the pattern “adverb/preposition + noun”: ˈbypass, ˈinside, ˈafterthought, ˈovercoat. e. Those nouns with the pattern “adjective + noun” (ˈblackboard, ˈgrandchild, ˈnobleman, ˈshorthand) with some exceptions (prime ˈminister, best-ˈseller, mobile ˈphone).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

272 f. g.

Beyond the Segment: Stress, Rhythm and Intonation

Those nouns resulting of the combination “verb + noun”: ˈbreakfast, ˈcookbook, ˈhangman, ˈpushchair. Those nouns resulting from the combination “verb + particle” or “particle + verb” (ˈhold-up, ˈincome, ˈmake-up, ˈoutset) with some excepctions (lie ˈdown).

(5.2) Compounds that have the primary stress on the second element, with usually a secondary stress on the first element The following types of nouns receive primary stress on the second element and secondary stress on the first. Note also that resulting compound adjectives and verbs also carry stress on the second element. a. Those nouns which include in the first element the ingredient or material that the second element is made of. Food items constitute a very numerous group, with the exception of compounds of –bread, -cake and -paste: apple ˈpie, gold ˈring, glass ˈjar, paper ˈbag (cf. ˈpaper clip). b. Those nouns which include some sort of location, positioning or time in the first element: (i) Location: a compound which has the name of a country, region or town as first element: Irish ˈcoffee, Spanish ˈomelette, Yorkshire ˈpudding, French ˈkiss. (ii) Compound place names generally: names of roads and areas within the town, including suburbs and districts and names of cities (Church ˈRoad, St Peter’s ˈSquare, Nelson’s ˈAvenue, Botany ˈBay), except those compounds ending in the actual word street (St ˈPeter’s Street). Also included in this category are the names of geographical and architectonic features, such as names of parks and gardens, as well as bridges, tunnels, buildings and stations (Abbey ˈWood, Castle ˈGreen, Bedford ˈPark, Tower ˈBridge, Clapham ˈJunction, Westminster ˈCathedral, Rose ˈCottage), including parts of these buildings (back ˈdoor, clock ˈtower, loo ˈwindow, front ˈgate, faculty ˈlibrary). Note, however, that compounds with room have stress on the first element (ˈbedroom), with some exceptions ( front ˈroom). (iii) Other compounds including location references with second-element stress are the names of publications, like newspapers and magazines, and associations, such as sports clubs or choirs (ˌSunday ˈExpress, ˌKent ˈMessenger, ˌCroydon ˈAthletic, ˌInverness ˈSociety). (iv) Positioning: right ˈwing, upper ˈclass, high ˈchurch. (v) Time: ˌChristmas ˈDay, ˌJanuary ˈsales, ˌevening ˈpress, night ˈbus, Old ˈEnglish. c. Nouns and adjectives with a numeral as the first element: first ˈclass, ˌsecond-ˈhand, third ˈworld, forty-ˈsix.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

273

d. Nouns with the structure “participial adjective + noun”: barbed ˈwire, split ˈsecond, ˌpromised ˈland, broken ˈpromise. e. Nouns of the pattern “-ing + noun” in which the -ing suggests a characteristic of the noun but does not indicate its function as above: leading ˈarticle is not an article for leading (cf. ˈsewing ˌmachine, or a machine that aids the activity of sewing), ˌdrinking ˈwater. f. Adjectives with the pattern “adjective + -ed/-ing form”, especially when used predicatively (long-ˈlived, big-ˈheaded, bad-ˈtempered, good-ˈlooking, hardˈworking, easy-ˈgoing). g. Nouns of the pattern “noun in -er/-ing + adverb/preposition”: passer-ˈby, washing-ˈup. h. Noun + adjective combinations: brand-ˈnew, snow-ˈwhite, world-ˈwide. i. Compounds which function as adverbs: North-ˈEast. j. Compounds which function as verbs, with an adverb as first element: down ˈgrade. (6) Difference between free phrases and compounds We have seen so far that most compounds take primary stress on the first element, especially in the case of an adjective as first element. In this case we are dealing with a real compound, a word which functions as one both semantically and grammatically. But sometimes these compounds may be confused with free phrases. Consider, greenhouse and green house: the former is a particular type of building made of glass for growing and protecting plants, whereas the second sequence refers to a house which is painted green. The difference between both sequences is signalled by accent only: the compound has primary accent on the first element, the free sequence on the second element, with secondary stress on the first. Phrases black ˈbird dark ˈroom green ˈhouse

Compounds ˈblackbird ˈdarkroom ˈgreenhouse

6.2.3 Prosodic stress As we have seen, prosodic stress is the stress given only to certain words in an utterance. The general tendency for English is that lexical or content words, which carry a lot of information, such as nouns, adjectives, adverbs and verbs, keep their primary stress, whereas grammatical or structural words, which carry a light information load, (articles, personal, reflexive and relative pronouns, possessive adjectives, prepositions, conjunctions and auxiliary verbs) lose it. Consider the following example, in which subscript L and G stand for

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

274

Beyond the Segment: Stress, Rhythm and Intonation

lexical and grammatical words, and where the general stress tendency is marked with ˈ:

AI 6.6 Prosodic stress GWe Gshall Ldisˈtinguish Lphoˈnetics G from L phoˈnology G for Lmethodoˈlogical Lˈreasons Gand Gbecause Git Lˈseems Lˈtrue Gthat Lphoˈnetics Gcan Gbe Lˈstudied Gwithout Lˈever Lˈreally Lˈgoing Ginto Lphoˈnology, Gwhile Lphoˈnology Gis Lˈclosely Ldeˈpendent Gon Lphoˈnetics Gfor Gthe Lˈdata Gon Gwhich Git Lreˈlies Gto Lpurˈsue Gits Lˈarguments.

However, exceptions to this overriding principle governing prosodic stress are observed in English, following rhythmical constraints, usage conventions and the intentions of the speaker. This means that certain lexical words may lose their stress whereas some structural words may be stressed, as we shall see (Finch and Ortiz Lira 1982: 115–116; Cruttenden 2001: 250; Collins and Mees 2003: 238–239). 6.2.3.1 Rhythmical variations Beat adjusts the number of stresses in an utterance, which in turn modifies the speed delivery of the unstressed syllables. Two adjacent full stresses is not a common beat pattern in English. This is why (1) In sequences of three content words, all of which could be stressed, the middle one tends to be unstressed, as long as it has no more than two syllables: a ˈcheap old ˈblanket, a ˈcheap tatty ˈblanket, a ˈcheap ordiˈnary ˈblanket. For the same reason, phrasal verb sequences may vary their stress patterns according to neighbouring stresses: a. Transitive phrasal verbs: ˈtake off your ˈcoat, ˈtake your ˈcoat off, ˈtake it ˈoff. b. Intransitive phrasal verbs: ˈsit ˈdown!, ˈdon’t sit ˈdown!, You may ˈsit down ˈhere. (2) In double-stressed words, whether simple or compound, the stress which is nearest another one in the utterance is lost or weakened: ˈcar-ˈpark: He went to the car-ˈpark, the ˈcar-park is full. ˈHeathˈrow: at ˈHeathrow ˈairport, the ˈtrain to Heathˈrow For the same reason, double-stressed adjectives, when used attributively, lose their primary stress and when used predicatively lose their secondary stress: ˌhand-ˈmade: a ˈhand-made ˈquilt, the ˈquilt is hand-ˈmade

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

275

(3) In some single-stressed words, a shift of stress may take place when used before another stressed word, for example, attributively: ˈalready: She’s ˈalready ˈleft, she’s ˈhome alˈready prinˈcess: a ˈletter from the prinˈcess, a ˈletter from ˈPrincess ˈAnne 6.2.3.2 Usage Certain parts of speech, actual words and phrases depart from the default stress pattern outlined above through ordinary usage. This is why: (1) wh-interrogatives, demonstrative adjectives and pronouns, possessive pronouns and negative anomalous finites are usually stressed. Note that exclamatory what, though, is usually unstressed, as in ˈWhy ˈcan’t ˈthat ˈdog of ˈyours ˈeat ˈthis? (2) The verb to be, even when used as a full verb, is usually unstressed, except in final position, as in She is ˈhere (vs. ˈHere she ˈis) or It was in the ˈhouse (vs. ˈThat’s ˈwhere it ˈwas). (3) Comment phrases are usually unstressed, as in “ˈCome on!”, he said, “Let’s ˈgo.” (4) Certain nouns when used in a very broad, general sense are usually unstressed (e.g. I ˈdon’t underˈstand ˈcertain things, I’ve ˈmet such a ˈlot of people). (5) The word street in street-names is usually unstressed, such as ˈLeicester Street (vs. ˈLeicester ˈSquare) or ˈKensington Street (vs. ˈKensington ˈGardens). 6.2.3.3 Emphatic and contrastive patterns Added meaning, highlighting emphasis or contrast, is achieved by altering the default stress pattern of an utterance. This is why: (1) Some grammatical words receive stress, as in He ˈphoned me (vs ˈHe ˈphoned me =it was him, not anybody else) or He ˈlived in ˈLondon (vs He ˈlived ˈin ˈLondon, not “near” London). (2) Anaphoric elements, which have just been repeated literally or referred to by a direct equivalent, are usually left unstressed, as shown in the examples included in (6) below: (6) a. A: You ˈdidn’t ˈtell me. B: No, I ˈknow I didn’t. b. They ˈvisited his ˈmother and ˈwhen they ˈleft his mother, they ˈcried. c. A: Would you ˈlike some ˈcoffee? B: No, ˈthanks, I ˈhate the stuff. Note that when the repeated elements in the utterance are used to express contrast they are usually both left unstressed, with the stress falling on the

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

276

Beyond the Segment: Stress, Rhythm and Intonation

contrasting words, as in You didn’t ˈtell me and I didn’t ˈask you. And if the anaphoric element changes function, then they are both stressed, as in They ˈvisited his ˈmother and ˈwhen ˈhis mother smiled, they ˈcried. 6.2.4 Nuclear stress Although we have seen that various words can carry stress in an utterance – regardless of whether or not they would carry it when pronounced in isolation – only one of those words carries what is known as the nuclear stress in the utterance. It is an observable fact of English that in connected speech, and particularly in what are called broad focus sentences, in which the most important element is generally presupposed, nuclear stress falls on the stressed syllable of the last important word of the utterance. In other words, the final stressed syllable of an utterance carries the nuclear stress. We will mark this typographically by highlighting it in bold. This is known as the nuclear stress rule. Consider example (7) below, in which the nuclear stress does not fall on the last syllable, but the last stressed syllable: (7) What would you like me to put it in? A common mistake by Spanish speakers is to place the nuclear stress on the accented syllable of the last word, whether it is an important word or not, that is, even if it is a grammatical word. For example, Spanish speakers will often pronounce *I went with her or *I saw him, with a stressed it or her. Once again, there are exceptions to this rule, so that the nuclear stress may be brought forward to an earlier position in the utterance, that is, it may be fronted. This fronting process is observable in: (1) Event or presentation utterances, particularly with intransitive verbs, or with transitive verbs followed by fairly predictable objects, as represented in (8): (8) a. He noticed the baby cried. vs The baby cried. b. He knew he drove the car. vs He drove the car. (2) Utterances ending in an adverbial: (9) I hate Mondays. vs She came to Manchester on Monday. (3) wh-utterances in which the wh-word functions as a modifier of a noun. (10) What did you take? vs What train did you take? Notice in this case that if the utterance is longer, then the nuclear stress falls in its default syllable: (11) What train did you take eventually?

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Stress and rhythm

277

Nuclear stress may also be displaced for reasons connected with narrow focus, with the intention of giving the utterance some special meaning. Thus, stress is displaced in the following contexts: (1) Whenever the last important word is obvious in the context of utterance. For example, when reading a story to a child and while pointing to the picture of a piglet in the book, we would not say This is a very naughty piglet, but This is a very naughty piglet. (2) For pragmatic reasons, distinguishing degrees of emphasis or contrast in sentences. Consider the utterances included in (12), which differ only in the placement of the nuclear stress: (12) a. I didn’t see him yesterday = ‘perhaps you did’ b. I didn’t see him yesterday = ‘I certainly did not’ c. I didn’t see him yesterday = ‘perhaps I spoke to him or emailed him’ d. I didn’t see him yesterday = ‘perhaps it was someone else I saw’ e. I didn’t see him yesterday = ‘perhaps it was the day before’ (3) Cases involving anaphoric reference, as seen before in the context of the distressing process of some words, are also affected by fronting of the nuclear stress, as in I would go with John but I don’t really like John. Now, anaphora may be presented in a very subtle way, and only the stress pattern will throw light on the possible pragmatic interpretations of such an utterance. Consider the example Jane said she’d been delighted long enough and Margaret offended her. The nuclear stress rule tells us that nuclear stress falls on the last stressed syllable, which seems to be the word offended. However, we also know that anaphoric references tend not to be stressed; so, if Jane’s comment that she’s d been delighted long enough is considered to be offensive in the context, then the word offended is an anaphoric reference, in which case it would not be stressed, which means it obviously does not carry the nuclear stress of the utterance, but this is displaced to an earlier position, namely, to the stressed syllable of the word Margaret, as shown in (13a) below. Now, if Jane’s comment is innocent and fair, then offended is introducing a completely new idea and is therefore eligible for stress, and because of its position in the utterance it would carry the nuclear stress, as it does in (13b). (13) a. Jane said she’d been delighted long enough and Margaret offended her. (Jane’s comment is offensive) b. Jane said she’d been delighted long enough and Margaret offended her. (Jane’s comment is fair)

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

278

Beyond the Segment: Stress, Rhythm and Intonation

6.3 Intonation We are all familiar with the fact that languages are not spoken in a monotone; in fact, one can easily recognise an automated, or robot-like voice, as in an onboard navigation satellite system or your mobile assistant, and in fact we tend to find it particularly annoying when someone speaks in a “monotonous” voice, an adjective which has acquired pejorative implications. What makes the difference between human and this automated kind of language is a question of pitch, which is what gives human language its melody. Pitch refers to the speed of vibration, more strictly speaking, the frequency of vibration, that is, the number of vocal cord movements per second (see § 1.2.2). We can distinguish, for example, four basic levels of pitch, 1 to 4 or Low, Mid, High and Extra High, respectively. When we speak, our voice travels on a line of declination, or downward scale, in which we introduce changes from one step to another, that is, rises and falls in the line of declination. This change of pitch level is basically what intonation is; however, to define intonation solely on the basis of this vocal fold activity would not render a complete picture, since together with pitch variation, other features such as loudness, rhythmicality and tempo make up the melody, or intonation, of human languages. Intonation is also a question of perception, as speakers are not always aware of producing it. Have you not experienced someone guessing where you come from on the basis of your accent, even if you are not aware of having one? However, having said that, some people are very good at producing voice impressions, taking into account various features, including idiosyncratic intonation (Tench 1996: 1–16; Collins and Mees 2003: 245; Roach 2005: 167–168; Wells 2006: 1–12; Cruttenden 2014: 299). Pitch variation has two distinctive functions. In the first place, pitch can help distinguish between different words which share the same phonemes; that is, tone, as it is known, is meaningful in that it distinguishes between the lexical, or “dictionary” meaning, and the grammatical, or inflectional meaning of words. This is the case of tone languages, in which each toneme is a lexically distinct variant of a given syllable so that they are phonetically distinguished from each other only by the tone of the vowel. Tonal languages are very common in Africa, the Far East and Central America, which means that as many as 70% of the world’s languages are tonal. This is, however, rarely the case in Europe and elsewhere in the Asiatic continent. Tonal languages may operate either on a register basis, that is, tones are distinguished by their pitch level relative to each other, which usually falls on monosyllables, or in the case of polysyllables on the whole word, as is the case of many African languages, or else on a contour basis, as is the case of the most widely-spoken tone language, Mandarin Chinese, in which each tone is made up of internal rising and falling

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

279

movements of pitch, so in the case of polysyllables, each syllable carries a different tone (Tench 1996: 3; Collins and Mees 2003: 246; Roach 167–169; Wells 2006: 4; Crystal 2008: 369–370). This brings us to the second function of pitch variation. If we think of English, we can easily observe how the word No, for example, may be produced with different intonation patterns under different circumstances, but its dictionary meaning of “negative reply”, for example, always stays the same. In most languages in Europe, this pitch movement, or intonation, plays a very important role since it can affect the way in which an utterance is understood in terms of speaker’s attitude, grammatical structure, organisation of the information, and intended illocutionary force, as we will see. Three related systems make up English intonation. These are: tonality, tonicity and tone. The first step in assigning prosody to a series of phonemes is to decide on the tonality: this has to do with chunking, or the splitting up of the spoken chain into manageable, meaningful units, or intonation groups. Once this is done, tonicity has to do with the stressing of the relevant syllable in each intonation group, that is, the nucleus of the group is located. Finally, tone has to do with pitch variations in the nucleus. These aspects of intonation all contribute together or in isolation to the correct delivery of any given utterance. Out of the several intonation marking methods possible, we propose one which is user-friendly while showing as much information as possible about the intonation patterns under analysis. By user-friendly, we mean that it is an in-text system with a reduced number of simple symbols which should make it easy to read from a script in spelling (which would therefore have no punctuation marks, which are the nearest orthographic equivalent of intonation), and to transcribe, both manually and in type, as follows:105 Table 44: Proposed intonation marking system Markup

Symbol

Further distinctions Symbol

tone boundary

|

end of utterance

||

Fall

\

high fall low fall

◝ ◟

Rise

/

high rise low rise

◜ ◞

fall-rise

\/

rise-fall

/\

ˈ

secondary accent

ˌ

primary accent nucleus onset

bold type ˚

105 Please note that no orthographic punctuation is used in the examples throughout this section.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

280

Beyond the Segment: Stress, Rhythm and Intonation

6.3.1 Intonation groups and Tonality In English, only a small number of particularly prominent syllables carry the tone in an utterance, which means that a unit larger than the syllable is needed in order to describe intonation patterns. So, we speak of intonation groups, or intonation or tone units, which are defined as the basic unit of speech melody, making up one complete pattern of intonation, usually lasting between one or two seconds. This chunking or breaking up of the spoken chain into separate intonation groups, which may or may not coincide with syntactical units, is known as tonality (Tench 1996: 31–49; Wells 2006: 187–205). Intonation groups are separated by tone boundaries, which are represented graphically by either a single bar (if there is a close connection between them) or a double bar (if not). Tone boundaries are not always easy to identify and it would, indeed, be simplistic to measure the length of an intonation group in terms of syllables. For practical purposes, let us say that, at the end of every clause there is an intonation boundary, and that, basically, a grammatical clause often represents one intonation group within the sentence: || Part of the sea wall in Lytham collapsed | and the railway line was badly damaged ||. This means that phonology, grammar and semantics are brought together by means of what is known as neutral tonality. Between coordinate clauses we usually place a boundary mark, as shown in (14), but usually not when there is an elided element, whether it be the subject or an object, which is the same for both clauses, as shown in (15): (14) || They played two more pieces | and the audience was delighted || (15) a. || They played two more pieces and received a standing ovation || b. || They played and sang two more pieces || Note that (15b) is potentially ambiguous. If an intonation boundary is placed after the first verb, as in (16), it is suggested that they played first (only instrumental) and then they sang two more pieces: (16) || They played | and sang two more pieces || However, other smaller grammatical units may represent an intonation group, as we often place intonation boundaries between phrases within a clause as well. We can talk of marked tonality when intonation units do not coincide, for several reasons, with whole clauses, as in the following cases (1) When the boundary between intonation units is placed between a long noun phrase and a verb phrase, as exemplified in (17): (17) || Wet and windy weather in the area | will continue until the end of February ||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

281

(2) When an intonational boundary is located within smaller phrases, such as (sentence) adverbs or adverbial phrases, either initially or in the middle of a clause, as shown in (18): (18) a. || Unfortunately | waves crashed over the harbour in Blackpool || b. || After all this heavy rain | a major salvage operation was launched || (3) When vocatives are used, particularly in initial position; if they occur in any other positions, they constitute the tail to what precedes them, as in (19): (19) a. || Kim | have you seen the news today || b. || Anne | as you know | is on sick leave || c. || Have you seen the news Kim || (4) Pseudo-cleft sentences involving what: (20) || What I saw | was his big friendly smile || (5) Parenthetic material in general, such as non-defining relative clauses: (21) || I met my brother | who arrived by train || (6) Also, question tags of the “checking” type, which are added to a clause in order to confirm the validity of the proposition expressed in the main clause, have their own separate intonation group, as in (22): (22) a. || They are not coming after all | are they || b. || She left all her luggage behind | didn’t she || (7) Question tags of the “copying” type, that is, showing an understanding of the value of the proposition, tend to be incorporated into the same intonation unit as the main clause, as it occurs in (23): (23) || The police found out who gave the false alarm did they || (8) Occasionally, and usually for reasons of emphasis or for some special effect, a tone boundary may be inserted between successive words in the same grammatical unit or even within one word: (24) a. || I | am not going || b. || To | mo | rrow || Tonality does vary according to whether the speech is being read aloud, and is therefore more carefully planned, or if it is spoken spontaneously. In the first case, intonation groups are usually longer and may therefore contain more stressed syllables than just the nuclear tone; whereas in unprepared speech the groups are shorter and usually have just one accent. Style and tonality are also closely related, since a formal pompous style will pause more often, and

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

282

Beyond the Segment: Stress, Rhythm and Intonation

have more stressed syllables overall, than ordinary speech, as shown in (25a) and (25b) below: (25) a. || Where have you \ been || b. || /Where | /have | /you | /been || 6.3.2 Structure of intonation groups and tonicity Every intonation group consists of a series of stressed and unstressed syllables. Typically, one of the stressed syllables carries the change in pitch on the declination scale we introduced earlier on, which makes it more prominent, in terms of longer duration and extra loudness. Consider (26) below, with these stressed syllables shown in bold type, others with just the usual stress mark: (26) || Desˈpite the ˈwarnings | some ˈresidents were not ˈleaving || The words warnings and leaving are the most prominent or salient words in each intonation group in this utterance, given the change in pitch on the syllables /wɔː/ and /liː/, with extra duration and loudness, however subtle this may be to the ear. This maximal prominence is what is referred to as tonicity, and the corresponding intonation pattern is called the nuclear tone. The most prominent syllable in an intonation group is the nucleus. Neutral tonicity tends to indicate a broad focus, i.e. all the information in the clause is new, and in this case the nucleus falls on the stressed syllable of the last lexical item in the group, i.e. at or near the end of the intonation group, and the function is to initiate pitch movement. This type of tonicity is also used when the focus is narrow but the new information is included at the end of the clause, which is its usual position anyway. Marked tonicity, on the other hand, happens when the nucleus falls on a lexical item which is not final, or on a grammatical item, and therefore signals this as new information, thus narrowing the focus. We find marked tonicity when expressing a contrast, both between lexical items or grammatical items, or even between morphemes, as in (27) to (30): (27) a. I would love a holiday by the sea. b. And I would hate a holiday by the sea. (28) Forget what they want and think of what we want. (29) He doesn’t live in Madrid but near Madrid. (30) The price of houses is not increasing, but decreasing.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

283

Marked tonicity is also employed in spoken discourse to avoid placing the nucleus of two subsequent intonation groups on the same actual lexical element, as when giving your phone number, for example: 5 386 746. Note that if the last digit of the second group was not 6, then the group would carry neutral tonicity, as in 386 749. Another context for marked tonicity to occur is in the presence of a final intransitive verb of movement or happening, which generally is rather empty of full lexical meaning, as contrasted in (31a) and (31b) below: (31) a. Helen’s going. b. Helen’s phoning. Nuclear tones show variation, as illustrated by the syllables /wɔː/ and /liː/ above, which do not seem to have the same pattern, as we shall see. Any syllables after the nucleus are known as the tail, and by definition are usually unstressed and continue the pitch movement that was initiated at that point. There may also be other syllables before the nuclear tone, which can be unstressed or weakly stressed. The first stressed syllable before the nucleus is known as the onset, and is graphically signalled by placing the symbol ˚ before it. The stretch between both syllables is the head of the intonation group. Any unstressed syllables preceding either the onset, or the nucleus in the absence of an onset, constitute the pre-head, i.e. an intonation group may have a prehead and no head, if there are no stressed syllables at all before the nucleus (Finch and Ortiz Lira 1982: 119–122; Tench 1996: 53–73; Cruttenden 2001: 256– 257; Collins and Mees 2003: 248–252; Wells 2006: 93–184; Roach 2009: 130–133). The general pattern of an intonation group, then, can be summarised as follows, with optional elements given in brackets: (pre-head) (onset) (head) nucleus (tail). So, going back to example (26) above, the analysis of its internal structure, represented in (32) below, gives us a more complete picture of each intonation group: (32) Des˚pite the /warnings | some ˚residents were not \leaving||106 Pre-head Onset Head Nucleus Tail

Des˚pite the /warnings des pite the war nings

some ˚residents were not \leaving some re sidents were not lea ving

106 If instead of ‘some residents were not leaving’ in the second intonation group, the sequence was ‘they were not leaving’, the stretch ‘they were not’ would be a pre-head and there would not be an onset, or stressed syllable before the nucleus, nor a head, i.e. no stressed syllables would precede the nucleus.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

284

Beyond the Segment: Stress, Rhythm and Intonation

6.3.3 Intonation patterns and tone The pitch movement initiated by the nucleus, i.e. the tone, or colloquially the tune, of intonation, is what is most often perceived by the ear, and what gives spoken utterances their melodic flavour. The pitch of the voice on the tonic syllable may move downwards or upwards, producing simple primary falling or rising patterns, or a combination of both in complex fall-rises, and the less common rise-falls. Let us have a look at these in turn, paying special attention to the intonation typically associated with statements, commands and questions, as well as the most usual intonation group combinations (Finch and Ortiz Lira 1982: 125–128; Tench 1996: 73–106; Collins and Mees 2003: 259–273; Wells 2006: 15–93; Roach 2005: 162–167; Cruttenden 2014: 278–280). 6.3.3.1 Simple tones: falls and rises These are prototypically associated with statements and commands in the first case, and questions in the other; however, the inherent connotations of these two tones may be used to modulate them, as we shall see.

AI 6.7 Falling tones

(1) Falls The most usual intonation pattern in RP English, a falling movement, always changes the melody to a low pitch, either from a high position, in which case we talk of a high fall, or else from mid pitch downwards, producing a low fall. The tail is low. A fall is considered a neutral tone, denoting a sense of conclusiveness, i.e. as if there is nothing else to follow. In the example below, if the utterance \No, with a falling intonation, is an answer to a previous question, it means that this negative reply is a final decision. The symbol used for a high fall is placed higher up just before the word which carries the nuclear tone, whereas the sign for a low fall is placed at the foot of the nucleus. Table 45: High tones Markup

High Fall

Low Fall

Nucleus only

◝ No

◟No

Nucleus last

I said ◝No

I said ◟No

Nucleus + Tail

◝ No more

◟No more

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

285

A falling tone is the commonest for statements, wh-questions and commands. Broadly speaking, a high fall makes them sound more lively and positive, with energetic connotations of interest and animation, whereas a low fall projects deliberation and distance, even coolness, as shown in (33): (33) a. \ Philip is coming tomorrow b. \ When is Philip coming c. \ Give me that chair Note that a yes/no question, which is usually pronounced with rising intonation, acquires the pragmatic force of a command when pronounced on a fall. This command will be more or less friendly depending on whether it is on a high fall or a low fall, as illustrated in (34): (34) a. / Will you ◜come here (question) b. \ Will you◝ come here (command; friendly) c. \ Will you ◟come here (command; distant) (2) Rises Rising patterns, a lot less frequent in RP English than falls, always work upwards to a mid or high pitch, either from a low position, known as low rise, which is more common, or from a mid pitch, or high rise. A rise is generally understood as a prompt for conversation to continue or to mean that there is something else to follow. In the example below, /No, with a rising intonation, can be used as a negative answer to a question and at the same time inviting more information on the part of the first speaker. The symbol used for a high rise is placed higher up just before the word which carries the nuclear tone, whereas the sign for a low fall is placed at the foot of the nucleus. Table 46: Rise tones Markup

Low Rise

High Rise

Nucleus

◞ No

◜No

Nucleus last

I said ◞ No

I said ◜No

Nucleus + Tail

◞ No more

◜No more

A low rising intonation is the commonest for yes/no questions, whereas a high rise signals surprise or doubt, and is also common in echo questions, i.e. repeating what has just been asked, as illustrated in (35):

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

286

Beyond the Segment: Stress, Rhythm and Intonation

(35) a. Do you like /◞ chocolate. b. Do you like / ◜chocolate. c. (Do you like chocolate) / Do I like ◜chocolate (thinking Hmmm. . .of course, I do!) Although normally pronounced on a falling pattern, as we have just seen, both wh-questions and commands when pronounced on a rising pattern convey a warm and gentle tone, which in the case of commands adds a sense of nonfinality. Compare the sequences in (36): (36) a. Now | go /◝ upstairs (high fall, as above) b. Now | go / ◟upstairs (low fall, as above) c. Now | go /◞ upstairs (low rise) Initial clauses, lists of items and parenthetical information are usually pronounced with a rising tone, given their non-final condition, as shown in (37): (37) a. After all the effort they / put into it the whole project was \ abandoned. b. Pack your / tent / pegs a / mallet a / torch a / sleeping bag and your \ boots. c. He’s so / attractive if you see what I / mean that I can’t \ help it 6.3.3.2 Complex tones: fall-rises and rise-falls Complex tones are not typically associated with particular grammatical or pragmatic categories; rather, their inherent connotations may be superimposed on any of them, as we can see below. (1) Fall-rise A fall-rise pattern shows movement from a high position down to low and up to mid pitch, with the fall on the nuclear syllable and the rest of the movement spreading over the tail, if there is one. It conveys a sense of hesitation or reservation, also implying doubt or partial agreement (as in e.g. I wouldn’t call it \/ perfect). This tone may be used to highlight two elements in an utterance, which is particularly useful both in wh- and yes/no questions, as the auxiliary verb would not normally be stressed, as shown in (38): (38) a. When \/ did he come? b. \/ Is he going?

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

287

(2) Rise-fall This unusual pattern moves up from mid to high pitch and down to low. If there is a tail, the movement can be completed in two different moves: either the rise falls on the accented syllable and the tail is produced at low pitch; or else, the accented syllable is produced at mid pitch which is then followed by a falling tail. The connotations of this pattern are those of the expression of strong feelings, such as surprise, doubt and (dis-)approval, adding a patronising effect, as in (39): (39) Oh I don’t know /\ whatever. For the sake of comparison, in the example below, \/ No, with a fall-rise pattern, is a negative answer which is expressing doubt or partial agreement, whereas /\ No, with a rise-fall pattern, may express strong surprise, for example, when listening to some unbelievable news. Table 47: Complex tones: Fall-rises and rise-falls Markup

Fall-rise

Rise-fall

Nucleus

\/ No

/\ No

Nucleus last

I said \/ No

I said /\ No

Nucleus + Tail

\/ No more

/\ No more

6.3.4 Functions of intonation So far we have seen what intonation is and how melody is made up of three interrelated systems – tonality, tonicity and tone-, which give it structure. Now we must turn to how intonation is used and what functions it plays. We are all familiar with the monotonous speech of answering machines, so it is obvious enough that intonation does something to the spoken chain when delivering a message. Most speakers seem to think that the only role of intonation is to express attitude, but this is not its only function. Four main functions of intonation can be highlighted, namely (a) attitudinal, (b) accentual or focusing, (c) grammatical, and (d) discourse and pragmatic. The attitudinal function has to do with the expression of speaker’s emotion and this is done by means of tone, and more specifically, by deviations from the default or neutral tones, which are normally associated with particular illocutionary forces. The accentual or focusing function is connected with the narrowing of focus on new information, but is also related to issues of emphasis and contrast. This is done by means of tonicity, in particular by shifting tonic stresses away from the default

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

288

Beyond the Segment: Stress, Rhythm and Intonation

positions. The grammatical function of intonation means that it helps disambiguate potentially ambiguous utterances, both at clausal and phrasal levels, by using tonality to alter intonation group boundaries as appropriate. Finally, intonation has an important role in the organisation of information in discourse, on the one hand, and in the encoding of the communicative functions of utterances on the other, for which all three intonational systems are at work (Tench 1996: 16–29; Collins and Mees 2003: 256–258; Wells 2006: 11–12; Roach 2005: 183–203; Cruttenden 2014: 284–298). Although intonational systems also operate in Spanish, and in some cases with some similarities, we cannot emphasise enough how important it is that intonation is properly taught to foreign students, since the “wrong” overall melody is more often a serious cause of linguistic misunderstanding, and even social misfortune, than the incorrect articulation of phonemes. 6.3.4.1 Attitudinal function Most speakers associate intonation with the expression of feelings and emotions, which are added to the semantic content of what is being said, to such an extent that attitude is a key factor in conversation – we have often heard “it’s not what you say, but the way you say it”. Intonation is a means of revealing attitude in speech, whether it be towards the listener, the actual message or the situational context, and it certainly gives away something about the speaker. However, it must be borne in mind that the expression of emotion in speech may be actually impossible, as in chorus community prayer, for example, or out of place, as in weather forecasting or, to a certain extent, in newsreading. Attitude is often expressed in colloquial, spontaneous conversation, but even in this context it is an optional component. Speakers have a variety of intonational resources which can be used to express emotion. Sequential elements, such as those aspects which are connected with tonicity and tonality, like shifting the nuclear stress, chunking the spoken chain into relevant units or measuring pauses, all contribute for a special effect. Prosodic elements include variations in loudness and speed, different voice qualities, and particularly pitch range, that is, changes in the width of the tones used. As we saw in Section 6.3.3. above, tones have neutral forms, which are mainly used to perform the grammatical, discursive and pragmatic functions of intonation, which we shall analyse below in more detail. It is mainly the complex tones and variations in the width of the pitch range of simple tones that are used to encode intonational attitude. (1) A fall-rise is associated with the expression of uncertainty, hesitation, and surprise. It may reveal the speaker as distant and detached or as if appealing to the listener’s reconsideration, with strict and stern undertones, showing reservation.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

289

(2) A rise-fall indicates the speaker is impressed, but also proud and conceited, as if mocking the listener, with a ring of dismissal. It is used for the expression of sarcasm and indignation. Compare the examples in (40): (40) a. ˅ what (fall-rise: reservation, surprise) b. ˄ what (rise-fall: arrogance, patronizing) (3) A high fall and a low rise are considered neutral tones, whereas a low fall and a high rise have an attitudinal function. A low fall is associated with boredom and resignation, disclosing a phlegmatic speaker, who is dispassionate, dull and detached; whereas a high rise suggests excitement, curiosity and eagerness, showing a speaker who seems to be concerned, enthusiastic or interested. Consider the examples in (41): (41) a. ◟what (low fall: boredom) vs ◝ what (high fall: unmarked, neutral) b. ◜what (high rise: interest) vs ◞ what (low rise: unmarked, neutral)

AI 6.8 Attitudinal function of intonation Finally, paralinguistic features, otherwise known as body language, are often revealing of speaker’s attitude. Facial expressions, like raising an eyebrow, and other body movements, such as shrugging a shoulder, play an attitudinal function and are an expression of emotion and the speaker’s inner state of mind, which may be accompanied by tears, laughter, coughing, and so on. Kinesics has proved most insightful in detecting lies in court, incriminating suspects, for example, in situations in which other forms of linguistic analysis fail to be useful. The attitudinal function of intonation is so highly valued by speakers that efforts are made to express this function in writing. Whereas tonality may be shown by the use of commas or question marks, and tonicity may rely on the use of bold and italic fonts, the orthographic expression of attitude is not an easy task. This may explain the recent development of a “dictionary” of emoticons107, a graphic representation of attitude in everyday conversations in texting and chatting contexts.108 Prosodic elements and, specially, facial expressions and vocal effects are reflected in these little faces, a metacommunicative

107 Emoticons have gone as far as not only encoding prosody, but almost becoming a new language altogether, to the extent that complete messages can be given by using only these icons – the advantage being the mutual intelligibility among speakers of different languages – at the expense of misunderstandings if icons are not interpreted correctly. 108 Before emoticons existed, efforts were made to convey attitude by means of ‘icons’ constructed on the basis of the combination of certain graphic characters available in an ordinary keyboard, such as :), which suggests a smile and possibly an associated rising tone.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

290

Beyond the Segment: Stress, Rhythm and Intonation

device which speakers eagerly use in order to make their attitudinal connotations clearer, and thus avoid the risk of misunderstandings, particularly in a medium which makes use of much abbreviated written language and hardly any punctuation, with messages overlapping in time and many times without a very clear context. Consider the following Whatsapp conversations:

Figure 50: Whatsapp Emoticons as a graphic representation of attitude

Conversation 1 is a mere transaction of information about the impossibility of going the next day. The only expression nearing emotion is encoded in the semantics of the Speaker B’s answer, ‘what a pity’, which could have been said with sarcasm, annoyance or impatience, but Speaker A does not know. His only reply is a factual postponement until Saturday, which could have been said on an inconclusive low rise, to which B closes with an emotionless farewell greeting. Conversation 2, instead, illustrates Speaker B showing emotion: the choice of an icon with the mouth curving downwards and a tear, denotes a high fall, expressing concern and worry. This clue is retrieved by Speaker A, who is then prompted to respond with a reassuring “Don’t worry!” before the inevitable postponement until Saturday, which is however suggested with a friendly and knowing icon, complete with a smile and a wink, suggesting probably a lively and definite high fall tone, which is the cue for the animated final goodbye. 6.3.4.2 Accentual function We saw in section 6.2.3. that nuclear stress may be displaced within a given utterance for various reasons. This prominent syllable we called the nucleus of the intonation group. This shifting of the nucleus will result in the highlighting of the information the speaker wishes to focus on. This is referred to as the accentual or focusing function of intonation and is related to tonicity. The nucleus of the intonation group may be used to signal a narrow focus on a piece of new information when everything else is given, or treated as given or implied – i.e. recoverable from the linguistic context, the conversational situation, common ground between speakers or general knowledge. In this case,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

291

marked tonicity functions to mark narrow focus on items which are not final lexical items (also see § 6.3.2. above). Consider, for example, the sequences in (42), where you can see how the focus of information changes as we shift the nucleus: (42) a. I only bought one carton of milk. b. I only bought one carton of milk. c. I only bought one carton of milk. d. I only bought one carton of milk. e. I only bought one carton of milk. f. I only bought one carton of milk. Other contexts of use in which intonation plays an accentual function are, as we have already seen, when contrast or emphasis is needed or the lexical content of the utterance requires it. Ambiguity can also be resolved by tonic shifting, as occurs in (43): (43) a. || He had ˈreasons to \ give || (= he wanted to give, with an implied direct object, which is implied, for example, money) b. || He had \ ˈreasons to give || (= he had some reasons that he wanted to give; reasons is the direct object of give in this case). In this case, as we shall see in the following subsections, tonicity often combines with tonality, if a new intonation group is created, and tones, which are added to the new groups, and these processes may result in the production of different syntactic structures, attitudes and even speech acts.

AI 6.9 Accentual function of intonation 6.3.4.3 Grammatical function This is mostly related to tonality: two structures which are parallel on the surface can be disambiguated by producing them with different intonation boundaries, which will have an effect on the syntactic function of some elements and the meaning of the overall utterance – a fact which is not always reflected orthographically. If a new intonation group is created, shifting of tonic stresses and the addition of new tones may have to be considered. Therefore, the three systems of intonation -tonicity, tonality and tone- often combine and have a role in this grammatical function of intonation. We will illustrate this by means of a selection of some of the most common cases. Please note that some of the examples could be dealt with under more than one heading. The following arrangement is for ease of reference.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

292

Beyond the Segment: Stress, Rhythm and Intonation

AI 6.10 Grammatical function of intonation (1) Vocatives Vocatives at the beginning of an utterance tend to have their own intonation group, are stressed and are said on a rising tone. Final vocatives, on the other hand, are usually said as the tail of the previous group. Consider the sequence, Mary my neighbour is coming tomorrow, which has two possible interpretations, illustrated in (44): (44) a. || Mary | my neighbour is coming tomorrow || b. || Mary my neighbour | is coming tomorrow || c. || Mary | my neighbour | is coming tomorrow || (44a) illustrates a case of an initial vocative: the speaker is addressing Mary to let her know that her neighbour, whose name we do not know, is coming the next day; in (44b) the speaker is saying – it is not known who he is talking to- that a neighbour of his called Mary – and not any other Mary he may knowis coming. That is, (44b) is a case of a defining use of the appositional element. In (44c), in turn, by placing an extra boundary after Mary, the appositional element becomes non-defining. Consider now (45), where the possible vocative comes at the end of the clause: (45) a. || Help | John || vocative b. || Help John || direct object Because final vocatives tend to be tails of the previous group, ambiguity rises as this element could be taken to be the direct object of the verb, as in (b). By placing an intonational boundary between both elements, help is shown to be a noun addressed to someone called John, whereas it is not known who the command to help John in (b) is addressed to. (2) Appositions Tonality is the key factor, as we have just seen, in disambiguating appositions from other structural units, as in, for instance, the sequence They called John a doctor, offering two possible appositive structures presented in (46): (46) a. || They called John | a doctor || b. || They called John a doctor || In (46a), the verb is a mono-transitive one, with John as the direct object and a doctor as an apposition to this, so that the meaning of the utterance is that ‘they called John, and John is a doctor’. In (46b), we have a case of a ditransitive verb, with John as the indirect object and ‘a doctor’ as the direct one, the meaning

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

293

being ‘they called a doctor for John’. There is a third meaning which cannot be distinguished intonationally: ‘They ascribed a doctorate to John’. (3) Relative clauses Tonality signals the defining or non-defining nature of relative clauses. For instance, a sentence like My friends who live in Lancaster arrived yesterday, could broadly show two different intonational patterns, illustrated in (47), which denote different syntactic structures: (47) a. || My /friends | who ˚live in /Lancaster | a˚rrived \ yesterday || b. || My ˚friends who live in /Lancaster | a˚rrived \ yesterday || The interpretation of (47a) is that of a non-defining relative clause, meaning that my friends arrived yesterday, and they all happen to live in Lancaster, implying by means of the appositional clause that all the friends I have are those who live in Lancaster. The interpretation offered in (47b) is that of a defining relative clause, functioning as a post-modifier of the head noun friends, denoting that it is only those friends of mine who live in Lancaster that arrived yesterday, but implying that I may have other friends elsewhere, and who have nothing to do with “arriving yesterday”. (4) Transitive versus intransitive verb complementation In the utterance, || They’re leaving Susan ||, the main verb, leaving, can be interpreted as transitive or intransitive: (48) a. || They’re leaving Susan || b. || They’re leaving | Susan || If the utterance is said as one single intonation group, as in (48a), then the verb is transitive and Susan is its direct object. However, as mentioned above, if Susan is a final vocative, in which case it is normally the tail of the group, then an extra intonational boundary has to be added for disambiguation purposes, as in (48b). Now consider an apparently similar example: (49) a. || They’re leaving all of them || b. || They’re leaving | all of them || Although (48) looks very similar to (49), syntactically, the function of the element all of them cannot be that of vocative, but it is a case of a postposed subject, as a kind of explanation. This means that it would be actually possible to find a context in which an utterance, given a particular selection of lexical items, posed a three-way distinction. Consider the following sequences: (50) a. They’re leaving my friends. b. She’s leaving Susan.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

294

Beyond the Segment: Stress, Rhythm and Intonation

Tonality can help disambiguate between the direct object function of the final noun phrases and the other two functions, but it is not enough to differentiate between the vocative and the postposed subject functions, with an intonational boundary before them in both cases. Tonicity does not help either, as both groups are very small and therefore carry the tonic stress on the same syllable regardless of the function. It is tone, in this case, which plays the crucial role. A falling tone in the first intonation group, denoting the finality of the statement, will suggest the vocative meaning, whereas a rising tone, denoting more information to come, will suggest the postposed subject meaning. Compare the sequences in (51): (51) a. || They’re \ leaving | \ my friends || versus They’re / leaving | \ my friends || b. || She’s \ leaving | \ Susan || versus She’s / leaving | \ Susan || (5) Phrasal verbs versus verb + prepositional phrase Phrasal verb constructions can create ambiguity, even if the particle is prepositional in both cases, as we can see in the following example: (52) a. || He looked | after my sister || b. || He looked after my sister || Although in both cases the sequence after my sister is a prepositional phrase, (52a) shows an intransitive use of the verb look with the PP as a time adverbial, which is shown by an intonation boundary between them, whereas in (52b) look is a prepositional verb with the PP as a verb complement. (6) Reflexive versus emphatic pronouns This is another case showing how tonality alone is not enough to disambiguate between two potentially ambiguous structures. The grammatical distinction between the structures (a) transitive verb + direct object (reflexive), as in I hurt myself, and (b) (intransitive) verb + subject complement (emphatic), I’ll send the letter myself, rely on tonicity, as the pronoun belongs in the same intonation group as the verb regardless of the function. In the transitive version it is the verb that carries the tonic syllable, but in the subject complement one, the reflexive pronoun, which could be given its own intonation group for further emphasis. Compare (53a), transitive, and (53bi) and (53bii), subject complement, the latter version, with a greater pause and two intonation groups, also adding the connotation of an afterthought. (53) a. || I have never taught myself || bi. || I have never taught myself || bii. || I have never taught | myself.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

295

(7) Double-verb constructions Some “verb + to-infinitive” constructions are potentially ambiguous, and placing a pause and creating and extra intonation group aids understanding. Consider the following example: (54) a. || She stopped to eat peanuts || b. || She stopped | to eat peanuts || The interpretation rendered by (54a), with one single intonation group, illustrates a transitive double-verb construction, with peanuts as the direct object, with the meaning ‘cease’, that is, she ‘ceased peanut eating’, equivalent to “she stopped eating peanuts”. (54b), however, shows an intransitive use of the verb stop and the sequence to eat peanuts as a purpose clause, that is, she stopped doing whatever it was she was doing at that moment, in order to eat peanuts. In fact, the two examples illustrate opposite meanings, as in (54a) she did not eat peanuts, but in (54b) she did. (8) Negative context In the context of a negative main clause followed by an adverbial clause of reason (introduced by because) or of result (introduced by so that), tonality helps resolve the potential ambiguity between parallel structures which, in fact, as in the previous subsection, denote opposite meanings. Consider (55) below: (55) a. || I didn’t buy because he advised me || b. || I didn’t buy | because he advised me || In (55a), the reason clause, embedded in the same intonation group as the main verb is functioning as an adjunct to the verb, and the implication of the whole sentence is that ‘I did buy’ but not as a result of following his advice, but for some other reason. In (55b), however, the reason clause is a sentence adjunct, or comment clause, and the implication is, in fact, that I didn’t buy and that was because he advised me not to. (9) Comment clauses and parenthetical information Something similar to the above is the case in the following examples, in which comment or parenthetical information is given their own intonation group to differentiate them from a transitive interpretation, as shown in (56): (56) a. || You know | I don’t like it || b. || You know I don’t like it || (56a) shows the discourse marker you know as a separate element, the main clause being “I don’t like it”, whereas (56b) shows the main clause “You know etc”, in which “I don’t like it” is a clausal direct object. Something similar happens in (57):

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

296

Beyond the Segment: Stress, Rhythm and Intonation

(57) a. || Their suggestion | to go by plane | was rejected || b. || Their suggestion to go by plane | was rejected || (57a) is a case of parenthetical information, the implication being that they only had one suggestion to make, which was “to go by plane”, and this was rejected. (57b) shows the sequence “to go by plane” as a noun modifier, which implies that they may have had other suggestions, but it was this particular one the one, and not the others, that was rejected. (10) Attachment of adverbials Adverbial mobility is a frequent source of ambiguity, and one which is very much exploited in humour. Appropriate tonality can help recover the meaning intended. Let us examine the following series of examples: (58) a. || Those who left | soon managed to get the lifts || b. || Those who left soon | managed to get the lifts || (59) a. || I was asking for the book I had seen | in the library || b. || I was asking for the book I had seen in the library || (60) a. || I saw the man | with the telescope || b. || I saw the man with the telescope || (61) a. || I saw the man | crossing the river || b. || I saw the man crossing the river || In example (58) the ambiguity is resolved by shifting the boundary to a position before or after the adverb, which indicates that the adverb is a modifier of the verb managed, in the main clause, in the (a) version, or of the verb left, in the relative clause, in the (b) version, with the consequent change in meaning: (a) means that the people that left managed to get a lift soon, whereas (b) means that only those who were quick enough to leave managed to get a lift. In contrast, examples (59) to (61) illustrate in the (a) versions adverbial expressions as modifiers of the immediately antecedent element, namely, “seen in the library”, “man with a telescope”, “man crossing the river”, whereas in the (b) versions, these adverbials belong higher up in the syntactic branching, that is, “asking in the library”, “saw with a telescope” and “ I crossing the river”. (11) Final adjuncts: manner versus comment Tonicity resolves the ambiguity of single intonation groups of the type: (62) a. || He fell unfortunately || b. || He fell unfortunately || c. || He fell | unfortunately ||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

297

(62a) illustrates the case of the adverbial of manner within the verb phrase, which carries the tonic syllable of the one single intonation unit, whereas in (62b), unfortunately is a comment adjunct, added to the whole clause; the tonic syllable of the intonation group falls on the verb. In (62c) this interpretation is further reinforced by giving the adverb its own intonation group, which means it would then also carry a nuclear stress. (12) Coordinated phrases Tonality is at work to disambiguate meanings in coordinated structures of different type, as represented in the following examples: (63) a. || I love chicken | risotto | and pie || (= I love three things) b. || I love chicken risotto | and pie || (= I love two things, but only the risotto has chicken in it). c. || I love chicken | risotto and pie || (= I love two things and both of them have chicken in them). (64) a. || The woman | and the girl in blue | are in my group || (only the girl is in blue; we don’t know about the woman). b. || The woman and the girl in blue | are in my group || (= both the girl and the woman are in blue). (65) a. || I washed | and polished the floor || (= washed is intransitive, whereas polished is transitive, i.e. the floor is the direct object of polished only; it is a case of two coordinated clauses). b. || I washed and polished the floor || (= both washed and polished are transitive, sharing the direct object the floor, that is, coordination takes level within the verb phrase and between the two heads). So far we have seen how intonation plays a role in the way we speak, which is superimposed on the actual message, to the extent that two similar messages can be delivered in different ways if we produce changes in the tonality, tonicity, and/or tone of an utterance. So, for example, if we consider an utterance like Take the keys, there are many different possible intonation units in which it could be “packed”, as shown in (66): (66) a. \ Take \ the keys b. Take the keys c. Take the keys d. Take the keys e. \ Take \ the keys

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

298

Beyond the Segment: Stress, Rhythm and Intonation

Tonality determines the number of intonation units that are made: for instance, one, as in (66b) to (66d), or two, as in (66a) and (66e). Tonicity allows us to shift the stress, which could fall on any of the three words, depending on the desired meaning. The three utterances in (66b) to (66d) above each illustrate a different shade of meaning of the utterance, which relies on shared ground between speakers. So (66b) means ‘take’ and not ‘send them by post’ for example; (66c) means ‘you know exactly which ones and what is special about them’; and (66d) suggests the keys as opposed to “a piece of wire, which could damage the lock”. In (66e) tonality and tonicity are combined, rendering an interpretation that combines those of (66b) and (66c), both carrying intonational prominence on being packed in corresponding intonation units. Tone will determine the attitude in which the command is uttered; as a fall, the difference lies on the variation of pitch, so that if said on a neutral high fall it sounds more lively and energetic than if said on a low fall, which shows signs of boredom and dullness, as in (67b), where (67c) shows that, again, the desired selection of tones can be combined with the intonation grouping intended and also the tonicity required: (67) a. ◝ Take the keys (= high fall: lively and cheerful) b. ◟ Take the keys (= low fall: boring and dull) c. ◟ Take \ the keys However sophisticated the system may seem to be so far, these are not all the possibilities that intonation allows, because we still have to consider at least one more major function, namely, the discourse and pragmatic functions. 6.3.4.4 Discourse function Discourse organisation consists in the chunking of information units into manageable pieces, deciding what elements count as old or given information in a message, and what count as new, and presenting all of this in some kind of order to the listener. Furthermore, conversation, in particular, is ruled by a series of norms that both speakers observe. Syntax and semantics tend to provide enough clues for discourse to be well organised, but intonation plays a very substantial role in these processes, either by supporting the existing layout or, most decisively, by disambiguating between otherwise equal utterances, where syntactic and semantic features alone fail to transmit the desired meaning, or by encoding an intended effect. The different prosodic elements of intonation often combine, as we have seen repeatedly in this section, but we can say that tonality has more to do with the first aspect of breaking up the spoken chain into information units; tonicity is related to questions of focus and old

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

299

versus new information; and tones support decisions concerning the staging of the information into major or leading units, and “minor” or “trailing” ones. Since the role of intonation in discourse is very much connected with other functions, such as the accentual and the grammatical functions already discussed, here we will take a brief look at some of the most relevant situations in which intonation plays this discourse function specifically.

AI 6.11 Discourse function of intonation

(1) Chunking Divisions into more than one intonation group of a sequence which would normally constitute one unit may obey a series of reasons, for example, the addition of an afterthought or the wish to reinforce a marked theme, as represented in (68): (68) a. || She is arriving tomorrow | by plane || b. || By plane | she is arriving tomorrow || Tonality also supports the addition of parenthetical information: || She is arriving | weather permitting | by plane || (2) New versus old information Neutral tonicity, as we have seen, means that the nuclear stress in an utterance falls on the tonic syllable of the last lexical item. If this item also represents new information, generally located at the end of an utterance, then the utterance is not marked. However, if new information is located elsewhere in the utterance, or is a grammatical element instead of a lexical item, then the stress is shifted accordingly, since we tend to stress new information, as illustrated in example (69): (69) || Did you meet Peter and Paul? No | I met Mark and Paul || (3) Broad versus narrow focus Neutral tonicity is the one we use to bring an utterance into broad focus, that is, we place the nuclear stress on the tonic syllable of the last lexical item, as in (70): (70) || She is arriving tomorrow by plane ||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

300

Beyond the Segment: Stress, Rhythm and Intonation

In addition to this, the linguistic system of English allows us a series of syntactic operations whereby only part of what is being said is brought into narrow focus, which is accordingly reinforced by a shift of the nuclear stress. Two such possibilities are represented in (71): (71) a. Who arrived first? || Ann I think || (fronting) b. || It was Ann that arrived first || (it-cleft) However, it may also happen that a narrow focus results from tonicity alone, as in (72), where the nuclear stress is unexpectedly shifted to initial position, instead of final position, because the final and medial slots are occupied by repeated material (Given information), and so the “unmarked”, or “neutral”, tonicity option is to place the focus on the initial constituent that carries New information: (72) Who arrived first? || Ann arrived first || (4) Staging of information We organise our discourse into chunks to which we give more or less importance, that is, some are dependent on others. The order in which these are given to the listener is irrespective of their importance. Tone helps signal this relationship. An independent chunk in an utterance tends to be said always on a fall, whereas dependent stretches are usually said on a non-fall, either a rise or a fall-rise, indicating that more information is to come – if they are preceding the independent or main element- or that they have just been added to the previous sequence and belong with it – when following it. When the dependent element comes first, the tone is usually a fall-rise, whereas if it comes after the main one, the tone is a rise, as shown in (73): (73) a. || There was complete silence when she \ left || b. || When she \/ left | there was complete \ silence || c. || There was complete \ silence | when she / left || If we compare (73a), the neutral version of this utterance, with the other two versions we will be able to notice the nuances in meaning that the different tones add. In (73b), the independent element is in the leading position and is therefore said on a fall-rise, whereas in (73c), where it comes after the main element, the tone is a rise. The function of the initial dependent fall-rise, preceding a falling tone, is to highlight a theme; if the theme is marked this will almost invariably have its own intonation group anyway, as shown in (74a) and (74b) below; but if theme

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

301

is not marked, then the choice of a fall-rise will not only highlight it but almost inevitably give it its own intonation group, as illustrated in (74c) and (74d): (74) a. || This / morning | my friend came || (= marked theme; neutral tone) b. || This \/ morning | my friend came || (= marked theme, marked tone) c. || My friend came this \ morning || (= unmarked theme; neutral tone) d. || My \/ friend | came this \ morning || (= unmarked theme; marked tone) Finally, we would like to mention again the question of lists. We have already seen that the usual tone between items in a list is a rise, and that the final element is said on a fall. This is the case of a closed list. However, if a list ends in a rise, this means the list is open and that there may be more information to come, even if it is not said. Something similar happens when giving choices, as shown in (75): (75) a. || We have va /nilla | or \ cream || b. || We have va /nilla | or / cream || It is from the tone alone that we know that the choice of icecream flavours in (75a) is limited to two, whereas in (75b) there are, or may be, more flavours, which have not been mentioned yet. The same takes place in alternative questions, represented in (76): in (76a) the question ends on a final fall, which implies the speaker is conceiving of only two possible places for the listener to stay, whereas the final rise of (76b) allows the listener to give a different answer. (76) a. || Are you staying in / London | or in \ Manchester || b. || Are you staying in / London | or in / Manchester || 6.3.4.5 Illocutionary function Intonation, particularly by means of the tone system, helps encode the illocutionary force of an utterance, so that it is delivered as a command, a question or a statement, for example. This is because there is not a one-to-one relationship between grammatical categories (declaratives, interrogatives, imperatives and verbless constructions) and illocutionary categories (statements, questions, commands and exclamations). This means that, although the interrogative mood usually encodes a question, and vice versa, a question is normally asked in the interrogative form, the linguistic system is flexible enough to allow for other combinations. Consider, for example the illocutionary category “command”, as

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

302

Beyond the Segment: Stress, Rhythm and Intonation

in (77a), which may be expressed by means of an interrogative, or the grammatical category “declarative”, as in (77b), which may actually encode a question: (77) a. Will you sign here, please? (interrogative = command) b. I wonder when she’s coming (declarative = question) Given this state of affairs, the grammatical structure of an utterance alone is not always indicative of its communicative function. Thus, it is easy to see now that intonation, and tones in particular, plays a crucial role not only in expressing the communicative function intended but, specially, in disambiguating between pragmatic categories which share the same surface grammatical structure, as can be seen in (78a) and (78b) (using example (66) above): (78) a. \ Take the keys (= command) b. / Take the keys (= question; suggestion) c. ◟Take / the keys (78c), on the other hand, shows that the discourse and illocutionary functions of intonation can also combine with the other three and therefore, as seen above, new meanings are possible: the first low fall command, denoting boredom and dullness, is followed by a question about the keys, which suggests ‘have you got them?’ or even ‘will you give them back to me?’; and, in any case, in this example, even more so than in the previous ones, the command to take is not necessarily related to the keys, as it could well refer to ‘taking something else’. In addition to the main four illocutionary functions that have been mentioned, there are others such as requesting, promising, threatening, complimenting, suggesting, offering, praying, greeting, cursing, insulting, and so on, the list is almost never-ending. All of them are brought together under the cover-term speech acts, which refer to what language is used for, given the fact that when we use language we are actually doing something with it. As an illustration, let us look at the following identical utterances, which are both said on a rising tone: (79) a. Can you / lend it to me? b. Can you \/ lend it to me? (fall-rise = request) The modulation allowed by the fall-rise helps define the illocutionary force of (79b) as a request rather than a simple question on a simple rising tone, as in (79a). Despite its richness, the tone system of the language is not sophisticated enough to enable each possible speech act to have its own characteristic melody. Hence, what the tone system does, by variations and combinations of

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

303

the basic falling and rising movements, is to allow for generalisations about the position the speaker adopts towards the listener along a cline from the power associated with falling tones to the deference related to the rising ones. In this section we will look at the main pragmatic categories, namely, statements, questions, commands and exclamations, and the role different tones play. We have already seen, in section 6.3.3. above, that the intonation patterns of English fall within two wide categories: falling and rising patterns, depending on the final movement in the nucleus of the intonation group. Falling tones tend to indicate finality, in the sense of conclusiveness, which is to say that they are used to disclose information the speaker has or to express his feelings, which, in other case, enables him to exert a certain sense of authority or dominance over the listener. Rising tones, on the other hand, indicate incompleteness, doubt and withholding of information, they are a prompt for conversation to continue and generally allow the speaker to show deference to the listener. Typically, statements, commands and exclamations are associated with falling tones, and questions with rising ones, as illustrated in the examples above, with departures from this signalling a different communicative function or modulating speaker’s attitude. (1) Statements a. The definitive fall. The neutral, unmarked or default tone for a statement is a fall, also known as a definitive fall, which means that what is said is a complete statement, definitive and final, which the speaker is expressing with confidence and resolution, as in (80) || I saw him \ in London || However, statements can be said on other tones, which have pragmatic connotations, that is, they add an extra layer of meaning which is not explicitly said, but implied. b. The fall-rise. A complete independent statement said on a fall-rise indicates that the speaker has some reservations about what he is saying, whether these are made explicit or not, that is, he is implying something else, which may or may not be left unsaid. Compare example (80) with (81), which has an implicational fall-rise: (81) || I saw him \/ in London || In this case, the reservation is not explicit, which means that it is left to the listener to resolve it: perhaps there was a “secret” reason for him to be in London, or else he should have been somewhere else, and so on. By using this tone, which is rising at the end, the speaker is somehow inviting the listener to either ask for what the reservation is or to offer it himself, depending on the degree of shared common ground between them:

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

304

Beyond the Segment: Stress, Rhythm and Intonation

(82) || I saw him \/ in London || What do you mean? I know, perhaps he was visiting Ann. This fall-rise can also be used for reasons of contrast, as in: (83) || I saw him in \/ London (= not in Manchester) One particular context of use of this fall-rise is in contrasting negative statements, whether the positive part is explicitly said or not. The above example would normally be said like this: (84) || I saw him \ in London | not in \/ Manchester A fall-rise is also used in tentative statements, in which the speaker does not wish to commit himself or is not completely sure about what he is saying. It also softens an otherwise more abrupt fall: (85) || How much money do you want? \/ Twenty pounds || This is the reason why it is also used in corrections which are made to sound more polite or friendly, as in (86a), and also in partial agreements, as in (86b): (86) a. So, you’re name is Mary \/ Susan || b. Did you enjoy the film? || The \/ music was good || (= not the rest of the film) c.

The rise. A rising tone in statements, known as the high raising terminal, uptalk or backpack, is a feature that entered the language in the 80s and spread quickly via the younger generations who, as is speculated, picked it up when travelling or watching foreign TV comedies. Because of the most usual meaning of a rise, typically used in yes/no questions, it sounds strange in a statement which is supposed to be final and conclusive, and a lot of people still find it annoying, as it is not quite clear whether the speaker is actually disclosing information, requesting it, being tentative or simply, most annoying, checking whether the listener knows what he is talking about. This leaves the listener in a position in which he does not always know whether he has to answer a possible question or just follow the conversation, as in: (87) A. Where’s the church? B. || Past / Sainsbury’s. . . .|| Speaker A is not sure whether Speaker B is actually checking whether he knows where Sainsbury’s is. The rise also suggests that something else is to follow, so Speaker A is waiting for more information, but if this is a case of uptalk, then, as far as speaker B is concerned, that was all there was to be said.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

305

A rise may be used in contradictions, but it sounds more unfriendly, as if on the defensive, than a definitive fall or the tentative fall-rise, as we have seen above: (88) You are late. a. || \ No || (= neutral) b. || \/ No || (= polite, deferential) c. || / No || (= brusque, protesting) Any of the tones above can be used not only for complete independent statements, but also in short answers to yes/no questions, for example, with all the nuances of meaning discussed above. Finally, we have already spoken about the lack of a one-to one relationship between grammatical and pragmatic categories. So far, we have seen declaratives functioning as statements, and vice versa, all the instances of statements in this section remain statements regardless of the tone used. However, a declarative on a rise is no longer a statement, but a question, and this is why we will deal with it in the following section: (89) || You are coming / tomorrow || (2) Questions There are questions of different types, and they do not all have the same tones. We will look at some of the most often used. (2.1) Wh-questions a. The definitive fall. Information questions, introduced by wh-words, are often pronounced on a fall, with a conclusive intonation. The reason for this lies in the fact that the presupposition enclosed in the proposition means that the speaker already knows at least part of the information, and therefore can use a more definitive tone: (90) || When are you \ coming || The final element is the presupposition, i.e. the fact that [you are coming] is being disclosed, and is consequently pronounced with falling intonation. b. The rise. A rising tone, whether it is a simple rise or a complex fall-rise, adds the overtones of deference and sympathy which are typically associated with it. This is known as the encouraging rise: (91) || When are you / coming || With this rising intonation, the question sounds much more friendly and welcoming than with the neutral default fall we have seen above.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

306

Beyond the Segment: Stress, Rhythm and Intonation

(2.2) Yes/No questions a. The yes/no rise. The neutral tone for a yes/no question, or polarity question, whether it is in the positive or in the negative, is a rise, known as the yes/no rise. The speaker is genuinely asking something for which he does not know an answer and is, therefore, encouraging the conversation to continue: (92) a. || Are you / coming || b. || Aren’t you / coming || This is also the case when the interrogative is actually encoding a request and not a question proper: (93) || Will you lend me your / car || b. The fall-rise. A fall-rise can be used make it even clearer that the interrogative is a request as opposed to a plain question in those cases in which there are no other illocutionary force indicators. Also, it sounds more persuasive than a simple rise. Contrast the following two examples: (94) a. Will you \/ lend it to me || (= requesting; persuasive) b. Will you / lend it to me || (= asking for a yes/no answer c.

The insistent fall is a yes/no question that is said on a fall. This makes the speaker sound more aggressive or formal: (95) || Are you \ coming || It is also used when the question has to be repeated if the listener has not heard properly, denoting some boredom or fastidiousness on the part of the speaker: (96) A. || Are you / coming || B. -Sorry, did you say something? A. Are you \ coming

(2.3) Declarative questions Note that, as mentioned above, there are cases in which it is the tone, rather than the grammatical form, that makes an utterance a question. So we can talk of declarative questions, which are also pronounced on a yes/no rise. The above example could be issued as: (97) || You are / coming || The reasons for choosing one grammatical form or the other may have to do with politeness issues -the declarative being perhaps less invasive- or with how much the speaker is assuming to be already true – the declarative implying that the speaker thinks he already knows the answer.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

307

(2.4) Echo questions These typically repeat all or part of the words of the previous speaker, whether these were a statement, a question of whatever type, or a command, in order to query them, ask for a repetition or to simply check understanding. Echo questions are said on a rise: (98)

A. I’m coming tomorrow| B. || Coming / tomorrow ||

(99)

A. When are you coming? B. || When am I / coming ||

(100) A. Are you coming tomorrow? B. || Am I / tomorrow || (101) A. Come tomorrow! B. || Come / tomorrow || This rise is known as the pardon-question rise and is also used whenever we want to enquire about any element in the previous utterance, either by using one of the “pardon” words, such as sorry or pardon, quoting the actual element that is being checked, or changing it into a wh-word, which may or may not be fronted: (102) || I’m coming tomorrow / Sorry || a. Coming / tomorrow b. Coming / when c. When are you / coming d. / You are coming Notice that if the query focuses narrowly on one of the elements of the previous utterance, there may be also a shift of stress. Examples (102a) to (102c) focus broadly on the previous utterance, questioning the whole of it, whereas in (102d) the speaker focuses on the element I only. (2.5) Tag questions Tag questions, variable reduced syntactic structures consisting of only an auxiliary verb and a pronoun, are usually added to the end of a statement in colloquial English. This is done on a polarity basis, in such a way that if the statement is positive, then the tag is negative, and vice versa, a negative statement is followed by a positive tag. This type is known as balanced or reverse polarity tag questions. More rarely, the tag and the statement may show the

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

308

Beyond the Segment: Stress, Rhythm and Intonation

same polarity, in which case we talk of unbalanced tag questions or copy tags. Consider the following examples: (103) a. Monica plays the flute, doesn’t she? (statement = positive; tag = negative) b. Martin can’t drive a car yet, can he? (statement = negative; tag = positive) c. Andrea will phone later, will she? (statement = positive; tag = positive) d. Angie doesn’t really like the project, doesn’t she? (statement = negative; tag = negative) Now, what is interesting about these tags is that their intonation is pragmatically determined, that is, tag questions may encode two different illocutionary forces -their grammatical form remaining the same-, and therefore act either as a question proper or else as a (request for) confirmation on the information which is given in the statement: in the first case, the speaker is not sure about the truth condition of the statement, whereas in the second, the speaker is only checking for confirmation of what he already suspects is true. Intonation, then, plays a disambiguating role, conveying the pragmatic force of the tag. As already discussed in Section 6.3.3.1, a rising tone is open-ended and denotes a true question, whereas a falling one, with its sense of finality, is the usual one in statements. Therefore, tags of the first type are said on a rise, and tags of the second type are said on a fall. Please note that the nucleus of the tag intonation group always falls on the verb and never on the pronoun, which is a common mistake among learners of English. Compare the following examples: (104) a. || He is not going back to Africa | / is he || = rising tone: true question b. || He is not going back to Africa | \ is he || = falling tone: request for confirmation The less usual unbalanced tags do not often have a separate intonation group, but if they do they are then said on a low rise. Consider the examples in (105): (105) a. || He is going back to \ Africa is he || (most usual intonation pattern) b. || He is going back to Africa | / is he || c. || He is not going back to Africa | / isn’t he || The above also applies to tags added to commands -as they show the same polarity-, which have the illocutionary force of a request, not a question. If pronounced on a rise, a positive tag softens the force of the command, whereas

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Intonation

309

a negative one, also on a rise, conveys a more demanding attitude on the part of the speaker, as in: (106) a. || Bring those boxes nearer to me | / will you || b. || Bring those boxes nearer to me | / won’t you ||

AI 6.12 Tag questions Resembling tag-questions, but actually uttered by a different speaker and with the same polarity as the previous one, elliptical short questions may be used as answers to a previous questions o statement, in order to keep a conversation going. The neutral tone for these is a yes/no rise, as in (107a) and (107b), whereas an insistent fall will denote hostility, scepticism or even contempt, as it occurs in (107c): (107) a. || I didn’t really like it / Didn’t you || b. || I have decided to stop it / Have you || c. || I have decided to stop it \ Have you || (3) Commands a. The fall. The neutral tone for a positive command is a fall, as in || Speak \ softly ||. b. The rise. The illocutionary force of a command may be softened by the use of rising intonation. It can also act as a reminder of a previously issued order or as gentle encouragement to get on with it, as in || Speak / softly ||. c. The fall-rise. In much the same way a fall-rise turned a question into a request, a command becomes a warning if said as in (108a), or it may be used in negative commands, as in (108b): (108) a. || Speak \/ softly | or I’ll call the po\lice || b. || Don’t speak so \/ loudly || (4) Exclamations a. The fall. The neutral tone for the expression of emotion in exclamations, interjections and greetings is the fall, regardless of their grammatical form, as shown in (109). Two common greetings are always said on a fall, namely Hi! and Cheers! (109) a. || How \ lovely || b. || \ Awesome || c. || Isn’t that \ good || d. || \ Thank you ||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

310

Beyond the Segment: Stress, Rhythm and Intonation

b. The rise. Exclamations and greetings on a rise denote a more friendly, approach-based attitude, and are also used as cues to encourage the continuation of the conversation, as exemplified in (110a) to (110d). Note, however, that, contrary to expectation, thank you on a rise denotes a routinary attitude, as shown in (110e). (110) a. He/llo || b. Good / bye || c. Good / evening || d. O/ K || e. Thank / you || 6.3.4.6 Other functions Intonation also plays other roles in our everyday life, such as aiding our memory – what is referred to as the psychological function- or signalling our social position. For example, it is often the case that people find it much easier to memorise phone numbers in a certain sequence but not in others. If you are used to saying your phone number in groups of three digits with their particular cadence you will find it more difficult when trying to give your number in pairs. Socially, it is also a well-known feature that newsreaders, for example, have a characteristic intonation, which makes them different from priests. Even within newsreaders, television and radio reporters, who rely solely on voice, sound different; also, whether reporting a happy item of news or a dramatic event, their voice, i.e. their intonation, is crucial for a professional delivery.

6.4 The prosody of English and Spanish compared and advice 6.4.1 Stress and rhythm English is a stress-timed language whereas Spanish is syllable-timed, as we saw above in section 6.2.1. Added to this, difficulties in word and sentence stress will cause the rhythm to be different in the productions of SSLE, who will tend to treat English as a syllable-timed language instead of giving it the stress-timed rhythmic distribution which is natural in native speech. Whereas Spanish word stress can be decided simply in relation to the syllables of words and their endings according to some straightforward rules, placement of stress in English is, indeed, a very complex issue. Although the learning of word accentual patterns is not as large a task as might be imagined considering that in conversations monosyllables may account for more than 80 per cent

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

The prosody of English and Spanish compared and advice

311

of words, the accent of polysyllabic words should be learnt when they are first acquired (Cruttenden 2014: 333). There is no simple way of predicting which syllable or syllables must be stressed. One obvious difference between the two languages is the fact that Spanish indicates stress through accent marks, whereas English does not. Therefore, whenever foreign speakers learn a new word, they must learn it individually with its stress. Incorrect placement of stress may be a major cause of intelligibility problems and of foreign diction. There are some tendencies in Spanish, which, if carried over into English, may lead to incorrect stress placement. Particular problems can appear with multisyllabic cognates, or “related words between two languages that are similar or identical in form and meaning” (Knorre 2009: 8), as illustrated in the examples given below extracted from Sedláčková (2010). 30–40% of all words in Spanish are cognates of English (Knorre 2009: 8). Table 48: Accentuation disparities between English and Spanish cognates Word class

Stress in English

Stress in Spanish

Noun

ˈanimal ˈdoctor ˈnation ˈoptimism ˈstudent voˈcabulary

aniˈmal docˈtor naˈción optiˈmismo estuˈdiante vocabuˈlario

Adjective

eˈffective inˈtelligent imˈportant ˈordinary ˈterrible

efecˈtivo inteliˈgente imporˈtante ordiˈnario teˈrrible

Verb

exˈport inˈvite inˈsist ˈoperate ˈsignify

exportar inviˈtar insistir opeˈrar signifiˈcar

Adverb

eˈxactly

exactaˈmente

As can be seen from Table 48, whereas in English stress is most commonly placed on the first syllable, in Spanish the two most frequent patterns are those in which the stress is placed either on the penultimate or on the last syllable. In addition, you can see that, unlike English, Spanish has a written accent mark when demanded by the Spanish accentuation rules, that is, in proparoxytone ‘esdrújulas’ words and oxytone ‘agudas’ words ending in vowels, and .

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

312

Beyond the Segment: Stress, Rhythm and Intonation

English prosodic stress falls on the last lexical word in the utterance, i.e. the word that carries the most relevant meaning, whereas Spanish systematically places it on the last stressed syllable of the sentence, regardless of whether this last word is lexical or grammatical. Therefore, a common mistake by Spanish learners is that represented in (111) below, where no emphasis on the stressed words is at all meant: (111) a. *I went with her b. *I love you Contrastive uses of sentence stress pose problems for Spanish learners and may give rise to pragmatic misunderstandings. This is because new information is typically presented after given information, receiving more stress, as it is considered to have more prominence in the sentence (and in the wider context of the conversation) than information already known to the listener. The different meanings expressed in English through sentence stress are conveyed in Spanish through other resources such as word order or the use of particular words. For example, while in English sentence stress serves to differentiate between Mary went to the shop (as an answer to what did Mary do?) and Mary went to the shop (as an answer to who went to the shop?), Spanish has different means to mark these contrasts, as in Fue a la tienda, with ellipsis of the subject in the first case, or Fue María¸ with subject-verb inversion in the second. The following points may serve as general guidelines in order to avoid interferences concerning stress and rhythm (Finch and Ortiz Lira 1982: 87–117; Estebas Vilaplana 2009: 138–181): (1) Every English word has a definite place for stress: it is most commonly placed on the first syllable but many words are stressed on the second syllable. Some words (mostly compounds) may have two stressed syllables. (2) The placement of stress is English depends on the grammatical category of the word. (3) The weight of the syllable (i.e. whether the vowel is short or long as well as the number of consonants that follow it) plays an important role as well. (4) Stress rules for simple words are mostly unpredictable and they are so complex that it seems easier to treat stress placement as an inseparable part of the word and learn it when the word itself is learnt. Rules for complex words’ stress placement are much easier than those for simple words (see § 6.2.2.2. above). Problems may arise with compound words. Because Spanish uses word order to show contrast, learners may be unfamiliar with the use of stress to mark the distinction between noun compounds and nouns with an adjective modifier, of the ‘blackbird and black ‘bird type. Spanish often stresses the second element of compound words, which

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

The prosody of English and Spanish compared and advice

(5)

(6)

(7)

(8)

313

means that speakers may incorrectly use the same stress pattern in English compounds. A most important aspect concerning English stress is the effect it has on the quality of vowels. Whereas Spanish maintains a strong vowel quality in both stressed and unstressed syllables, in English vowel weakening occurs, as already noted, so that unstressed syllables are usually produced with a weak vowel [ə i u ɪ ʊ], although some strong vowels can also appear in unstressed syllables. As an illustration, consider the word potato /pə ˈteɪtəʊ/, where the two unstressed syllables contain weak [ə] and strong [əʊ]. Vowels can also be omitted in unstressed syllables giving place to syllabic consonants, as already explained in Section 1.3.3.3). SSLE should pay particular attention to vowel weakening in order to improve their English pronunciation, as well as to prevent communication problems because in some contexts vowel weakening has a contrastive effect. Consider, for example, the difference between exercise /ˈeksəsaɪz/ and violet /ˈvaɪələt/, on the one hand, with vowel weakening to /ə/ of the first and second syllables, respectively, and exorcise /ˈeksɔːˌsaɪz/ and violate /ˈvaɪəleɪt/, on the other, where no such weakening occurs (Fudge 1984). While in Spanish polysyllabic words have only one stressed syllable, in English especially long words tend to have more than one stressed syllables which may have either a primary or a secondary stress. One frequent mistake for SSLE is to produce English double (understand /ˌʌndəˈstænd/) or multiple (misrepresent /ˌmɪsˌreprɪˈzent/) stressed words with one single stress, which results in a rather strange foreign accent. In English there exist a number of (normally disyllabic) words whose grammatical category (verb or noun/adjective) is distinguished by the location of the stress (suspect /səˈspekt/ (v.) vs. /ˈsʌspekt/ (n)). SSLE should avoid producing such words with the same stress distribution, whatever their grammatical category, usually by placing the stress on the first syllable [ˈsaspekt]. We have already seen that in English phrases and compounds, which may be written differently (as two separate words, as two words separated by a hyphen or as one word), can be distinguished by the location of the stress: the latter tend to receive only (one) primary stress on the first element (greenhouse /ˈɡriːnhaʊs/), while phrases have two stresses, a pre-primary stress on the first word and a primary stress on the last word if it is the head (green house /ˌɡriːn ˈhaʊs/. One common mistake for SSLE is to pronounce both sequences with a phrasal stress pattern, that is, with two stresses. This may lead to obvious misunderstandings. In the case in point a listener may interpret that the referent is ‘a house painted in green’ when what is actually intended is ‘a place that is full of plants and vegetables’.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

314

Beyond the Segment: Stress, Rhythm and Intonation

Let us now consider rhythm. We have explained that English has stresstimed rhythm, that is, stress occurs at approximately equal intervals so that the duration between stressed syllables tend to be approximately the same, regardless of the number of intervening unstressed syllables. In contrast, Spanish has a syllable-timed rhythm, in which syllables are produced at more or less regular intervals so that that the duration of each syllable is fairly similar, irrespective of whether it is stressed or unstressed. Consequently, for a speaker of Spanish sentences like ˈMary could ˈsee and ˈMary could have been ˈseen differ in duration, that is, the higher the number of syllables, the longer the sentence. For an English speaker, in contrast, both last approximately the same amount of time since what matters in English is the number of stresses, not the number of syllables. Bearing this in mind, if SSLE want to sound natural when speaking in English particular attention should be paid to the following: (1) English stress-timed rhythm should be reproduced by reducing the duration of the unstressed syllables and producing (or omitting) weak vowels in unstressed vowels, as well as the weak forms of function words, instead of pronouncing syllables with similar duration without changing the quality of unstressed vowels. A good way to start acquiring English rhythm is through singing songs and reciting nursery rhymes. The tempo used can be marked by clapping. If the tempo is quicker, then more unstressed syllables should be produced between stresses to be able to keep the same rhythm. (2) Another aspect derived from English stress-timed rhythm is that when a phrase consists of three content (mainly mono- or disyllabic) words occurring in a sequence, the one in the middle normally loses the stress (little young boy /ˌlɪtl ̩ jʌŋ ˈbɔɪ/) to contribute to keeping a more even rhythmic pattern. Instead, SSLE tend to stress either just the word in the middle [lɪtl ̩ ˈjʌŋ bɔɪ] or each of the three words [ˈlɪtl ̩ ˈjʌŋ ˈbɔɪ], which sounds awkward to the native ear. (3) A further consequence of English stress-timed rhythm, stress-shift occurs in English but not in Spanish. It is the phenomenon whereby the stress of a word is shifted to the front to avoid the production of two consecutive or close stressed syllables. Thus, words like Japanese /ˌdʒæpəˈniːz/, which has two stressed syllables, one with a secondary stress (the first) and another one with a primary stress (the third), if followed by another word which starts with a stressed syllable such as noodles, the primary stress will be shifted to the syllable that originally carries the secondary stress ( Japanese noodles /ˈdʒæpəniːz ˈnuːdl ̩z/). In contrast, SSLE normally pronounce such sequences without stress shift ( Japanese noodles [ʤapaˈniːz ˈnuːdl ̩z]), which again causes a strong foreign accent. To minimise this effect, the application of the stress-shift rule is recommended.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

The prosody of English and Spanish compared and advice

315

6.4.2 Intonation Intonation patterns employed by English speakers differ from those of Spanish speakers, as we have seen, mainly regarding the much wider range of pitch that English uses: higher highs and lower lows are reached, and more often. As a result, Spanish may be perceived as rather monotonous and not so melodic as English. This way, the English speaker may get the wrong impression that the Spanish speaker lacks involvement or interest in the conversation; but most importantly, since intonation plays a series of linguistic functions, wrong melodic movement may lead to the conveyance of a wrong message. Some areas which are particularly problematic for Spanish speakers are: pitch range, since Spanish speakers seem to use too narrow a pitch range; final falling pitch movement, which may not sound low enough; and the rise-fall pattern, which is difficult for learners to produce, especially on short phrases or monosyllables (Sedláčková 2010: 42–46; 74–76; Gutiérrez 2005; Gutiérrez and Conde 1990; Finch and Ortiz Lira 1982: 56). In both languages, the two main melodies are based on rising and falling tones and combinations thereof; but, whereas in English, utterances begin at an overall high pitch, in Spanish utterances begin at a lower one; and while in English there is a large pitch change on the most emphasised word, in Spanish, the major pitch change usually takes place on the first stressed syllable. In practical terms, this means that in an English utterance such as I am going to the shop, the higher pitch used at the beginning would gradually lower until the end of the utterance, where the word shop would be given the most emphasis. If the same sentence was pronounced with a Spanish melody, most probably it would begin at a low pitch, which would then be raised on the first stressed syllable of the sentence, i.e. going. As far as the intonation contours for statements and questions are concerned, English and Spanish differ substantially, which means that interferences may occur. The following areas are problematic: (1) Statements Both English and Spanish use a final falling pitch for statements. But in clauses with two intonation units English frequently uses the fall-rise pattern, which is not the case in Spanish. Compare the following: (112) a. || Todavía no ha llegado el \ tren || ai. || The train has not \ arrived yet || b. || Si encuentras las llaves \ dímelo || bi. || If you find the \/ keys | let me \ know ||

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

316

Beyond the Segment: Stress, Rhythm and Intonation

(2) Interrogatives English and Spanish intonation contours for interrogatives depend on the type of question. For present purposes, we distinguish between yes/no questions and wh-questions. (2.1) Yes/no questions English yes/no questions end in a rising pitch when the speaker does not know the answer to the question: (113) a. || Is he / handsome || b. || Do you want some / coffee || In addition to the rising tone pattern, English also employs a falling tone if the speaker expects confirmation (e.g. in some tag questions), or if the speaker is sure that the answer to the question is positive: (114) a. || The weather is \ good | \ isn’t it || b. || Is he \ coming || In contrast, Spanish yes/no questions always rise at the end: (115) Es el libro / interesante || ‘Is the book interesting?’ The pitch is high on the first stressed syllable, then the contour line remains the same until the last stressed syllable where it rises again. Therefore, Spanish learners are expected to face problems maintaining the correct (falling) intonation pattern in English yes/no questions when the answer is expected to be positive. (2.2) Wh-questions Wh-questions in Spanish have a tendency to rise on the final syllable, whereas they are pronounced on a falling tone in English. This is a cause for frequent mispronunciation. (2.3) Tag questions Since in Spanish tag questions are always pronounced with a rising intonation, SSLE generally use a rise in English tag questions, regardless of their function, when in English, as already noted, they are generally uttered with a falling contour, if asking for agreement, or with a rising intonation, if asking for information. SSLE should hence become aware that if used for different purposes, as information- or confirmation-seeking devices, tag questions in English should have correspondingly different intonation patterns, rising in the first case or falling in the second.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

The prosody of English and Spanish compared and advice

317

Although the correct acquisition of the attitudinal functions of English intonation is one of the most difficult areas for SSLE, some effort should be made to master at least some of its uses, in particular those of fall-rise in, for instance, warnings, contradictions and reservations. Also relevant for learners aiming at a native-like RP is the tonal quality of this accent, which mainly derives from the types of pre-nuclear patterns that are used, most frequently involving a series of glides-down. A series of glides-up will sound comically enthusiastic, and it should be avoided in long sequences of syllables on a low level, which will be perceived as bored or even surly. SSLE should therefore make every effort to employ the tonality, tonicity and tone (pitch) range used by native English speakers in order to avoid giving the overall impression that their speech is monotonous or otherwise being “accused” of not showing enough enthusiasm or interest. Consider, for instance, thanks-giving exchanges such as A: −Thank you. B: −Thank you. By choosing a contrastive accentuation pattern focusing on you speaker B wants to emphasise that it is him/her, not A, who must be grateful. In order to succeed in such a challenging task, some training may be necessary to exercise the learners’ acting abilities, as well as their empathy and imagination (Cruttenden 2014: 335; Sedláčková 2010: 75; Estebas Vilaplana 2009: 182–234).

Further reading Most manuals for learners of English phonetics and phonology tend to include one or two chapters on stress, rhythm and intonation. Our experience tells us that this topic is still the “poor relation” in a general phonetics and phonology course for foreign students. This is not necessarily for lack of information on the topic: quite to the contrary, it is often the case that too much detail is given on certain aspects – such as the morphological structure of words in order to get accentuation right, or the intricacies of the internal structure of the English tone unit, for example – which make this an even more difficult and complicated area of practice for foreign learners. So, if you wish to go into further details along these lines, please refer to Gimson (1984, Chapters 9 and 10), Cruttenden (2014, Chapters 10 and 11) and Collins and Mees (2003: 123–132), (2009: 226–284), the latter two offering many interesting and practical exercises on stress and rhythm, and intonation. Roach (2005) offers an extensive and accessible analysis of English prosody, which has served as inspiration for this chapter: Chapter 10 focuses on the nature, levels and placement of stress within the word, whereas complex word stress is dealt with in Chapter 11, and intonation is extensively treated in Chapters 15 to 19.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

318

Beyond the Segment: Stress, Rhythm and Intonation

For a thorough study following Halliday’s functional linguistic treatment of intonation, the reader is referred to Tench (1996) and Wells (2006). In their aim at describing intonation objectively and with precision, both authors provide a descriptive framework of intonation as a system composed of tonality, tonicity and tone, known as “the three T’s”, which we have also adopted here. Based on these parameters, they give special attention to the crucial role intonation plays in expressing meaning, providing clear explanations of some of the key functions of intonation, namely the attitudinal, grammatical and discourse functions, as well as what we have labelled pragmatic. They both give many illustrative examples throughout, and an extensive collection of useful exercises, both written and recorded. Finally, descriptions of English prosody in contrast with that of Spanish or addressed to SSLE can be found in Monroy Casas (2012, Chapters 3, 4 and 5, 1980, 1981), Mott (2011, Chapters 7 to 10), Estebas Vilaplana (2009, Chapters 5 to 8), Gutiérrez (2004a, b, 2005), Gutiérrez and Conde (1990), Finch and Ortiz Lira (1982, Chapters 9, 11 and 12), and Sánchez Benedicto (1980, Chapter 5), to mention but a few.

Exercises 1.

2.

Choose six pairs of words from Table 43 in Chapter 6. Write two sentences to illustrate their different stress patterns. Mark them and then read the sentences aloud carefully. Example: insult I don’t like ˈinsults vs She inˈsults everyone when she’s angry Explain the stress shift processes in the following words. Group those that show the same pattern: kingdom

3.

demonstrative magnetic

journalist

idiomatic evidential

Place the stress on the following words with suffixes; find the (possible) correlate root: carriage memorial delicacy

advantageous alcoholic original

lemonade domain widen

cinematography glorify dogmatic

4. Place primary and secondary stresses on the following compounds; justify your answer: low classes Christmas holiday kitchen window grandson

front door fifty-four lemon juice guesswork

onlooker St Therese Cathedral Ivory Coast greenhouse

London Bridge pencil sharpener blackmail downpour

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

The prosody of English and Spanish compared and advice

5.

Stress all the words that carry prosodic stress in the following utterances, using the labels L for lexical words, and G for grammatical words, commenting on the exceptions that you find. 1. 2. 3. 4. 5. 6.

6.

The man who came and asked for her sounded rather foreign. That thick book on the lower shelf is not mine. The child himself had told me before. What lovely weather but how cold it is! Who was the man that was talking to you before the class? She said that their car would not start again.

Stress the relevant syllables in the following constructions with phrasal verbs: 1. 2. 3. 4. 5. 6.

7.

319

bring down the box / bring the box down / bring it down it blew up / it then blew up / it blew up suddenly give back the book / give the book back / give it back fill in the form / fill the form in / fill it in she dressed up / she couldn’t dress up / she dressed up nicely he gave up / he won’t give up / he gave up abruptly

Provide your own examples to illustrate the stress modifications the following words suffer because of rhythmic constraints: shop-window, first-floor, afterlife.

8. Place prosodic stress on all the possible syllables in the following utterances. Explain your choices. 1. 2. 3. 4. 5. 6. 9.

Don’t let down your brother, she said, and pay him back. I didn’t know where you’ve been! What a dark room! I can´t see a thing. The old blue car broke down in Queen Street. He offered her some cake but she hates sweet things. Nick went to the supermarket.

Place the nuclear stress in the following utterances, explaining the exceptions. 1. 2. 3. 4. 5. 6. 7. 8. 9.

Put it in the box. Put it in the box near it. Put it in the box (complaining that he was putting it outside the box) Go with him! The actors appeared. What did you buy? What detergent did you buy? What detergent did you buy yesterday? The pictures, he said, were extraordinary.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

320

Beyond the Segment: Stress, Rhythm and Intonation

10. The train was very late (about ten minutes). 11. The train was very late (over an hour). 12. I saw her. 10. Place the nuclear stress on all the possible words in the following utterance and explain the differences in meaning: Ruth gave me only one book 11. Explain the pragmatic implications – i.e. implicit meaning – produced by the possibility of shifting the nuclear stress in the following utterance: 1. 2.

The boy raised his fist and the teacher threatened him. Carol told the actor how well he had performed and then he praised her.

12. Analyse the internal structure of the intonation groups in the following utterances, on the basis of the structure given in Chapter 6: (pre-head) (onset) (head) nucleus (tail) 1. 2. 3. 4. 5. 6.

I don’t know I don’t know her I don’t know her name Who do you live with? Where do you live? Do you live in London?

13. Add the various optional elements (pre-head, onset, head and tail) to the following intonation groups, which consist of a nucleus only, in a variety of combinations, e.g. pre-head + nucleus; pre-head + nucleus + tail, or nucleus + tail, and so on. Mark up the new group using the appropriate symbols. 1. John 2. take

3. now

4. green 5. two

14. Consider these two conversations. Describe the possible intonation pattern of the utterances “What a pity”, “Perhaps on Saturday” and “I will try on Saturday” using graphic notation, and paying special attention to the use of emoticons. Justify your decisions.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

The prosody of English and Spanish compared and advice

321

15. Explain the difference between both answers to the question: Do you know about Susan’s new job? Answer A: /No. Answer B: \No. 16. Spanish is described as a syllable-timed language, whereas English is stresstimed, as we have seen in Chapter 6. This can make it difficult for Spanish speakers to sound natural in English, and vice versa. Pronounce the following sentences with Spanish syllable-timing first, and then with the correct English stress-timing pronunciation. 1.

Tropical storms have been migrating northwards, a paper in Nature says. 2. The Facebook Translation Department will make Bolivian a language option. 3. More than 500 museums, galleries and historic spaces across France are keeping their doors open. 4. In September 2012 a man in his forties was found dead in Shetland Avenue a suburban street in West Nottingham.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:13 PM

Chapter 7

7 Predicting Pronunciation from Spelling (and Vice Versa) 7.1 Introduction The main aim in this chapter is to help SSLE to increase the probability of correctly guessing the pronunciation of words that they read but are not familiar with. The problem is not, as some people say, that there are no rules relating spelling with pronunciation, but rather that there are very many of them, as well as very many exceptions, and also that surprisingly few people seem to know or understand the rules. This applies not only to speech-to-text rules (regarding how to spell, which will also be briefly dealt with at the end of the present chapter), but to text-to-speech ones as well (how to pronounce). The more rules one knows, and the better one understands and applies them, the more words one will find to have a pronunciation that is predictable. And it is of course important to avoid applying the text-to-speech rules that apply to Spanish but not to English. For example, the spelling almost never represents the sound /aʊ/, and never represents /θ/. It is perhaps especially in the case of vowels that text-to-speech is complicated. For this reason, we will begin with a description of how the vowelspelling system works (Section 7.2) and this will be followed by the spelling-tosound correspondences of consonants (Section 7.3). In particular, we will see how it seems less chaotic if one relates it to syllable division (the syllabic structure of words). We will deal with stressed vowels first, and leave unstressed ones until later. But firstly, brief comment is needed about the way vowels are classified. If we are studying speech sounds in isolation, a classification should be based on phonetic criteria, but if we want to understand how the orthography functions, and to know the spelling patterns that apply to different groups of sounds, then a classification should be in terms of such groups. The chapter closes with a summary of the sound-to-spelling correspondences of English sounds (Section 7.4) because they have already been described and exemplified in more detail in Chapters 3 and 4, in the Spelling subsection of each sound entry.

7.2 Spelling-to-sound correspondences of vowels This section describes the spelling-to-sound correspondences of stressed (Section 7.2.1) and unstressed (Section 7.2.2) vowels. This is followed by a discussion of silent V letters (Section 7.2.3).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of vowels

323

7.2.1 Stressed vowels The spelling-to-sound correspondences of stressed vowels are described in this section considering seven parameters: (1) syllable division (Section 7.2.1.1), (2) the contrast between “heavy” and “r-tense” vowels (Section 7.2.1.2), (3) vowels following /w/ (Section 7.2.1.3), (4) occurrence of lax instead of tense vowels (Section 7.2.1.4), (5) occurrence of tense instead of lax vowels (Section 7.2.1.5), (6) the contrast between lax and “heavy” vowels (Section 7.2.1.6), and (7) occurrence in digraphs (Section 7.2.1.7). 7.2.1.1 Syllable division As already explained in Section 1.3.3, both in spelling and in pronunciation, a syllable ending in a consonant (C) is described as closed, and a syllable ending in a vowel (V) is described as open. Let us consider for the moment that, as a basic rule, the V in a closed syllable is a lax one (/æ e ɪ ɒ ʌ ʊ/), as in the word plan, while the V in an open syllable is a tense one (/eɪ iː aɪ ǝu juː uː/), as in no, as illustrated in Table 49. All lax Vs are short, pure ones, except that /æ/ is long in some words if followed by certain voiced Cs. All tense Vs are long pure ones, or else diphthongs. Table 49: Vowels in closed and open syllables Example

Type of syllable: open / closed

Type of V

Pronunciation

plan

closed

lax

/plæn/

no

open

tense

/nǝu/

Now, plane is pronounced as one syllable (as final is almost never pronounced), but orthographically the word is to be considered as consisting of two syllables and as having the syllabic structure . So, although the spoken word is /pleɪn/, which is a closed syllable, the first written syllable (spelt ) is open, and it is this detail of the spelling that tells us that the V is tense and that the word is therefore /pleɪn/, as shown in Table 50. In fact, this is of course one of the main functions (although not the only one) of final mute , and one might say that the orthography, curiously, seems to obey the principle of syllable division more consistently than the phonology does. Table 50: Orthographic syllable division Example

Orthographic syllable division

1st syllable: open/closed

Type of V

Pronunciation

plane

pla.ne

open

tense

/pleɪn/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

324

Predicting Pronunciation from Spelling (and Vice Versa)

The above assumes that a C is to be considered as part of the same syllable as a following V letter (if there is one), whether a spoken counterpart of the second written V is pronounced or not. It should also be made clear that if there are two or more C letters, then as many of these that can begin a word begin a syllable. For example, many words begin with , so waste has the structure , which indicates the pronunciation /weɪst/. Silent is not the only V letter that shows that a previous one very probably represents a tense V sound, as shown in Table 51: Table 51: Tense vowels in non-final syllables Example

Orthographic syllable division

1st syllable: open/closed

Type of V

Pronunciation

planing

pla.ning

open

tense

/ˈpleɪnɪŋ/

wasting

wa.sting

open

tense

/ˈweɪstɪŋ/

pastry

pa.stry

open

tense

/ˈpeɪstrɪ/

On the other hand, no word or syllable ever begins (neither in spelling nor in pronunciation) with , which is only one of various C clusters that do not begin a word or syllable, nor with a double C, which is usually one C written twice (), as well as a few other letter combinations (). So in both cases the syllable boundary “splits” the C cluster, leaving the earlier syllable closed, which implies a lax V, as displayed in Table 52, therein lies the difference between such words as backing = bac.king /ˈbækɪŋ/ and baking = ba.king /ˈbeikɪŋ/. Table 52: Lax vowels in non-final syllables Example

Orthographic syllable division

1st syllable: open/closed

Type of V

Pronunciation

bandit

ban.dit

closed

lax

/ˈbændɪt/

planning

plan.ning

closed

lax

/ˈplænɪŋ/

7.2.1.2 The letter and other V sounds: “Heavy” and “r-tense” is the usual spelling of /ɑː/. is basically a C letter, which (whether it represents a /r/ sound or not) “closes” an orthographic syllable in many words like star, star.ring, star.red, just as other C letters close syllables that include a lax V (plan, plan.ning, plan.ned). Start, like pant, is also a closed syllable, and in star.ting, like in pan.ting, the C cluster is split, again leaving the earlier syllable closed (since neither words nor syllables ever begin with or ).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of vowels

325

/ɑː/, /ɜː/ and /ɔː/ are what will be called here a heavy V (to avoid the ambiguous word “long”). Certain other V sounds are spelt with a V letter followed by , but where the syllable is open, as in sta.re, sta.ring, sta.red (cf. pla.ne, pla.ning, pla.ned), in which case the sounds /eǝ ɪǝ ɑɪǝ jʊǝ/ will be called r-tense. This label reflects the fact that is part of the spellings of words in which such sounds occur, although they are represented, just like tense Vs, in orthographically open syllables, which mark them as r-tense instead of heavy. The above examples of words with heavy and r-tense Vs are presented again in the following table: Table 53: Heavy and r-tense vowels Example

Orthographic syllable division

1st syllable: open/closed

Type of V

Pronunciation

star

star.

closed

heavy

/stɑː/

starring

star.ring

closed

heavy

/ˈstɑːrɪŋ/

start

start.

closed

heavy

/stɑːt/

starting

star.ting

closed

heavy

/ˈstɑːtɪŋ/

stare

sta.re

open

r-tense

/steǝ/

staring

sta.ring

open

r-tense

/ˈsteǝrɪŋ/

So, in short, usually represents /æ/ or /eɪ/ depending on whether the orthographic syllable is closed or open, while usually represents /ɑː/ or /eǝ/, again depending on the type of orthographic syllable. The pattern is similar with the other Vs, as is reflected in Table 54, where the following variations should be clarified: (1) Variations regarding sounds in the column: a. Before dark l [ɫ] (but in no other cases), the tense V [ǝu] (spelt ) is replaced by [ɒʊ] (the diphthong in Galician words like cousa ‘thing’), as in cold, pole, if it is not followed by a V sound in the same morpheme. So it can be said that the spelling of hole and hope have different pronunciations that are different allophonic realizations ([ɒʊ] [ǝu]) of the same phoneme /ǝu/. b. The letter sequences , and represent [ɔːl]. However, in the latter case, tends to represent [ɔːl], if is not followed by a V, in words that are Germanic and/or monosyllabic; but it represents approximately [æ] or [aː], if it is followed by [ɫ], in a few non-Germanic words such as banal, canal, algebra, and just syllabic /l ̩/, when is word-final and unstressed as in final.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

326

(2)

Predicting Pronunciation from Spelling (and Vice Versa)

c. Note that the words for, fort (with “heavy” spellings) and fore, story (with “r-tense” spellings) all have the same [ɔː] sound. Presumably, previously different sounds have merged, in most idiolects. Variations regarding the column: a. There are two lax sounds often spelt with : over twenty roots (i.e. excluding derivatives) have the V sound /ʊ/, and over fifty have /ʌ/. b. Over fifty of spellings are pronounced /ʌ/. c. There are also two tense sounds spelt with : /juː/ and /uː/. But /juː/ does not occur after /r/, nor after /C l / (i.e. l preceded by another C).

Table 54: Vowels of the four groups Type of orthographic syllable

Type of V sound

Spellings with vowel letters



/



closed (not by )

lax

æ plan planning

e pet petty

ɪ pin pinned

ɒ hop hopping

ʌ or ʊ cut put cutter putting

open

tense

eɪ plane planing

iː Pete Peter

aɪ pine try final fined

ǝu (or [ɒʊ]) hope hole

juː or uː cute rude

closed by

heavy

ɑː star start starry tarry

ɜː her referring

ɜː stir stirred

ɔː for fort boring

ɜː fur furry

open, but follows V letter

r-tense

eǝ stare staring

ɪǝ revere adherent

aɪǝ tire tyre spiral tired

fore story

jʊǝ pure fury

AI 7.1 Pronunciation of the letter The above observations and tables may be said to represent the primary or most basic text-to-speech rules for vowels. Of course, how to pronounce words that we read is not always so simple, as there are various types of “departures” from the above (or exceptions to them), as indicated below. But in many cases,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of vowels

327

exceptions to a rule have something in common which is the basis of a secondary rule or “sub-rule”. In other cases deviation is arbitrary and not predictable by rule. 7.2.1.3 Different V sounds after /w/ Although the lax Vs spelt and are usually /æ/ and /ɒ/ respectively and the heavy Vs spelt and are usually /ɑː/ and /ɔː/, they are different after /w/, in (world) and (quarter) spellings. Table 55: Sounds after RP /w/ Spelling

Pronunciation

Examples

Exceptions

a

ɒ

wand, wander, what, quantity, warrior

/æ/ before a velar: wax, wag(gon), swang

o

ʌ

won, wonder, worry (also one wʌn/)

wobble

ar

ɔː

war, warm, quarter

or

ɜː

work

(s)wore, (s)worn.

7.2.1.4 Lax instead of tense The following lists a number of cases in which a lax V is pronounced where the spelling suggests tense V (see also § 7.2.3): (1) The word natural, derived from nature (with /eɪ/), has lax /æ/ instead of /eɪ/, even though the is not doubled. This violation of the syllable division principle is due to trisyllabic laxing (i.e. laxing of a tense V in the antepenultimate syllable of suffixed words), a pronunciation change which occurred after spellings were established. Other word-pairs like nature – natural include: serene – serenity, fable – fabulous, Bible – biblical. The same phenomenon has occurred in many penultimate syllables, especially where -ic has been added to a monosyllable, as in cone – conic (also in numerous words spelt with a digraph, e.g. please – pleasant), as well as in some monosyllabic derived words, especially the past forms of several irregular verbs (dream – dreamt). In such cases one might call the phenomenon disyllabic laxing and monosyllabic laxing respectively (see § 7.2.7 for a discussion of other examples of laxing). Note that in words ending in , the penultimate syllable is nearly always stressed and includes a lax V even if the word is mono-morphemic (comic), or is a derivative but from a word without any tense V (atomic).

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

328

Predicting Pronunciation from Spelling (and Vice Versa)

(2) Other written words wrongly suggest tense Vs because of graphotactic constraints, i.e. spelling “rules” that serve no purpose and simply mislead readers. The letter and certain others are virtually never doubled, so for example /lɪvɪŋ/ is misleadingly spelt instead of *. (3) C letters are almost never doubled after a V digraph, but single as in looking ( would be a less ambiguous spelling.) A few exceptions are foreign words that have kept their original spellings, e.g. chauffeur (cf. Spanish chofer). (4) Doubling is also missing in some words with irregular V spellings, e.g. coming, money. But notice, however, that the in worry, as well as the /ʌ/ sound spelt can be considered regular after . 7.2.1.5 Tense instead of lax In the following cases a tense V is pronounced where the spelling suggests lax V: (1) The spellings of words like find seem, at least at first sight, to wrongly imply a lax V in so far as only one V letter is written. On the other hand, addition of final (like in plane) would not make the word conform to spelling rules either, since the syllable division would be which implies a lax V like in the word bandit. By contrast, a few word-spellings like most and post, which end in C clusters that can be initial, would indeed be less misleading with added (cf. waste with the syllable division wa . ste). On the other hand, it seems that all the common English words spelt with final are pronounced with /aɪ/ (except wind when meaning ‘moving air’), so the learning effort is negligible. (2) Note also that word-final is pronounced /iːz/ instead of /ɪz/ when it is the plural ending of Latin words ending in (crisis, crises). (3) It was mentioned above that is virtually never doubled. To add to the confusion, it is never word-final either, so /lɪv/ is spelt instead of *. One cannot therefore guess purely from spellings whether the V before a is lax, as in have and giving, or tense, as in save and arriving. 7.2.1.6 Lax vs. heavy Although is used to spell heavy and r-tense Vs, it is also of course used to spell the C sound /r/, which can follow a lax V. How can one know that in preferring the penultimate V is /ɜː/ but in herring it is /e/? Quite simply, preferring is the gerund of prefer, in which clearly represents /ɜː/ since no words end in /er/ in RP (SSBE), and addition of an ending such as -ing does not change the V sound in the root. However, there is no verb her, and herring is in fact a mono-morphemic noun and the V is lax. Tarry as an adjective derived from tar

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of vowels

329

is /tɑːrɪ/ but as a mono-morphemic verb is /tærɪ/ (cf. also fur # ry hurry ┤).109 Accordingly, the relevant part of Table 54 could be extended as follows: Table 56: Vowels of the four groups (extended) closed by

heavy or lax

ɑː star.ry tar.ry æ ├marry┤

ɜː prefer.ring

ɜː stir.red

ɔː bo.ring

ɜː fur.ry

e ├herring┤

ɪ ├mirror┤

ɒ ├horrid┤

ʌ ├hurry┤

├tarry┤

7.2.1.7 Digraphs Vs are usually tense, whether the syllable is open or closed, if a V digraph (two consecutive V letters) is written. For example, represents /eɪ/ as in rain, and /ǝu/ as in road. Digraphs are in fact an alternative way of indicating that a V is tense, or r-tense as in stair if a digraph is followed by . Although this complicates the orthography, an advantage is that homophonous words are distinguished in writing, as in plane – plain and stare – stair, thus reducing confusion for readers. There are many exceptions to the rule that a V spelt with a digraph is tense. is usually /iː/ as in please, but it is often /e/. However, most examples of the latter are cases of disyllabic laxing as in pleasant, breakfast, (derivatives of please, break) or monosyllabic laxing as in dreamt, or else words in which is followed by or (bread, breath, feather). As a result, there exist some homographs (lead, read) much in the same way that the pronunciation of some digraph cannot be guessed, as in the case of , which may have a pronunciation with lax /ʊ/ (soot, hood) or one with tense /uː/ (boot, food). A few other letter sequences function as V digraphs even though they do not consist of two V letters. is a frequent spelling of /juː/, of /ǝu/, and of /aɪ/ ( few, know, sigh). Note that is another spelling of /ɔː/, even though this sound is not a tense one but a heavy one.110 There are two diphthongs that are always spelt not according to syllable division but with digraphs: /aʊ/ and /ɔɪ/. The former is spelt or , and the latter or . Regarding text-to-speech, there is a problem in that often represents /ǝu/ (as mentioned above), and often represents /uː/ or /ʌ/ as in group and cousin. 109 The symbol # here represents a boundary between morphemes, while the special brackets ├ , ┤ are a convenient way of indicating no morpheme boundary, i.e. a monomorphemic word. 110 Non-finally, is usually written instead of , and so compare pause with paw.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

330

Predicting Pronunciation from Spelling (and Vice Versa)

Word-final V digraphs virtually always represent tense V sounds (assuming final stress). Their use is in general somewhat redundant, since lax Vs do not occur in stressed word-final position, so a single V letter would be understood as probably representing a tense V anyway. To conclude this section on stressed vowels, Tables 57111 to 59 list the main stressed sounds of each of the common digraphs (and some less common and/ or longer spellings) when not followed by , as opposed to the pronunciations of digraphs followed by that are registered in Table 52 above.

Table 57: V letter followed by other V letter or Spellings

Pronunciation

Examples

Comment

More examples

ae

iː e

formulae haemorrhage

(esp if Latin pl. end)

anaemia

ai, ay

eɪ

plain, play

au w

ɔː ɔː

cause law

(not usu bf V:) (/ǝw/ away)

ea

iː e

meat head

freq bf

Exceptions

/e/ said /iː/ quay /ɑː/ aunt, laugh please pleasure, feather

/ɪǝ/ idea, really

ee

iː

meet

ei, ey

iː

receive, key

eu

juː

feud(al)

ew

juː uː

dew grew

/j/ omitted af /r, ʧ/ or C+l

ie

aɪ iː

lie believe

Syl. final, if only letter bf non-final

oa

ǝu

road

/ɔː/ in (a)broad

oe

ǝu

toe

/uː/ in shoe

oi, oy

ɔɪ

noise, toy

/iː/ or /aɪ/ (n)either /aɪ/ height, /eɪ/ weigh(t), they /e/ leisure pseudofew, chew, blew sewer

/ǝu/ sew /ɪ/ sieve

111 In this chapter the following abbreviations are used: af = after/following, bf = before/ followed by, freq = frequently, and usu = usually.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

331

Spelling-to-sound correspondences of vowels

Table 57: (continued) Spellings

Pronunciation

Examples

oo

uː

food

ou

aʊ uː ʌ

noun group cousin

ow

aʊ ǝu

now know

ue

juː uː

due rude

uy

aɪ

buy

ye

aɪ

Comment

More examples

Exceptions /ʌ/ blood, flood /ʊ/ hood

(see also ough below)

county, south soup, youth country, couple, southern, young bow, brown bow, own

/j/ omitted af /r/ or C+l

/ɒ/ knowledge

true, blue, clue guy

final

rye

good-bye, dye

Exceptions in foreign words not fully anglicised ai oi

aɪ wa

balalaika bourgeois

Russian French Rare digraphs

ao

eɪ

gaol /ʤeɪl/ also spelt jail chaos /ˈkeɪɒs/, where represents /eɪ/ + /ɒ/

eo

iː ǝu

people yeoman

Table 58: V followed by (V)C(C) Spell.

Pron.

Examples

Comment

More examples

al

ɑː

almond

bf in same morpheme

balm

igh

ɑɪ

sigh, sight

augh

ɔː ɑːf

daughter laugh(ter)

ough

ɔː uː ǝu aʊ ʌf ɒf ʌp

ought through (al)though plough enough cough hiccough

ɒʊɫ

shoulder

oul

Exceptions

caught draught only before

/ɑʊ/ drought dough rough, tough

also spelt /ʊ/ could

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

332

Predicting Pronunciation from Spelling (and Vice Versa)

Table 59: Digraphs + Spelling

Pronunciation

Examples

air

eǝ

Air, airie

ear

ɪǝ eǝ ɜː ɑː

ear, year, tear (n) bear, tear (v) early, earth, pearl, search heart(y), hearth

eer

ɪǝ

beer, career, -eer mountaineer

ier

ɪǝ

pier, frontier

oar

ɔː

oar

oir

waɪǝ

choir (homophone of )

oor

ɔː

door

our

ɔː aʊ ɜː

four, mourn our journey, journal(ism/ist), adjourn

7.2.2 Unstressed vowels For help with deciding which syllables are stressed and which are not, see Chapters 5 and 6. Here we shall briefly consider how to guess the correct V sound in an unstressed syllable, based on its spelling. SSLE will already realise that the reduced or weak V /ǝ/ is the most usual sound but not the only one. As a general rule, the vowel sound in an unstressed syllable depends on what the sound would be if the syllable were stressed, and this is often indicated by the spelling, and/or by the pronunciation of a related word with a different stress pattern. E.g. the in suppose shows that the would, if stressed, be the short vowel /ʌ/ – as it indeed is (although not fully stressed) in the word supposition – and the weak (unstressed) version of the short vowels /æ, ɒ, ʌ / (and often /e/) is /ǝ/. /ɑː ɔː ɜː/ also tend to weaken to /ǝ/ (cf. confirm vs confirmation). But there is a tendency for /ǝu ju: uː/ to be shortened without being weakened to /ǝ/. Hence the in argue and argument would probably be /juː/ if stressed (as suggested by the spelling), but if it is not stressed, then it is a shortened /ju/ sound rather than /ǝ/. Similarly, /huː/ (who) is weakened to /hu/, not /hǝ/. The sound /ɪ/, apart from being one of the (stressed) lax Vs, is also another weak V which for example “replaces” /eɪ/ in the unstressed ending spelt ,

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of consonants

333

as in village. It is also a very frequent alternative to schwa when the V spelling in an unstressed syllable is , as in believe, which is pronounced either /bǝliːv/ or /bɪliːv/. 7.2.3 Silent V letters It was mentioned earlier that word-final is nearly always silent (i.e. it does not itself represent a sound), except in words with no other V letter (e.g. he /hiː/, the /ðǝ ðiː/) and in a group of words mostly from Greek (and often similar in Spanish), as in apostrophe /əˈpɒstrəfi/, where the V in the previous syllable is neither stressed nor tense. Elsewhere, final acts as a marker indicating that: (1) the previous V is tense ( fate /feɪt/, face /feɪs/) or r-tense if the intervening C is silent ( fare /feə/); (2) pre-final represents /s/ instead of /k/, irrespective of whether the previous V is tense or lax (ice /aɪs/, office /ˈɒfɪs/); (3) pre-final represents /ʤ/ instead of /g/ (rage /reɪdʒ/). Likewise, in regular past forms and some plural forms, non-final is usually written but silent (washed /wɒʃt/, clothes /klǝuðz/). However, in some cases a silent is added to show that a word ending in single is not being used as a plural noun (or 3rd person verb ending), as in tense /tens/, in which case the is voiceless. Finally, an unstressed V in the second of three syllables is very often silent, the word being pronounced as two syllables instead of three, as in chocolate /ˈʧɒklɪt/, business /ˈbɪznɪs/, general /ˈʤenrl ̩/.

7.3 Spelling-to-sound correspondences of consonants This sections explains the spelling-to-sound correspondences of RP consonants, taking into account the parameters of voicing (Section 7.3.1.1), as well as the effect of “silent” C letters (Section 7.3.1.2). After this, a summary of these tendencies is presented (Section 7.3.2), in addition to other pronunciation details (Section 7.3.3). 7.3.1 Voicing and “silent” C letters 7.3.1.1 Voicing As regards voicing, special attention should be paid to the following: (1) always represents an unvoiced sound (/k/ or /s/, or very rarely /ʧ/). (2) is pronounced /ks/ (unvoiced), unless it is immediately followed by a stressed vowel sound in the same word, in which case it is voiced /ɡz/. (Cf. /ɪkˈspekt, ˈtæksi/ vs /ɪɡˈzæm/)

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

334

Predicting Pronunciation from Spelling (and Vice Versa)

(3) in final is nearly always unvoiced (course). (4) between vowel letters is usually (but not always) voiced (close when verb but not when adjective). As already noted in Chapters 4 and 5, those voiced consonants that have unvoiced counterparts (i.e. /b d g v ð z ʒ ʤ/) are not usually completely voiced at the end of words, but only what we might call semi-voiced. So, to help our listener avoid confusion between, for example, card and cart, or ride and write, we pronounce the preceding vowel (if it is a supposedly long one) in a shorter manner than usual. In this way, there are two small differences between the words, instead of only one, the difference in length being often more audible than the difference in voicing.

AI 7.2 Pronunciation of the letter 7.3.1.2 Silent C letters Silent letters in any given word are those that do not themselves represent any sound. Besides, graphotactic constraints, whereby in some words a letter is written, or else omitted, for no valid purpose so that , for example, is virtually never doubled but always followed by a V letter (pronounced or silent),112 SSLE should be aware of the following points concerning silent consonants (C) in RP: (1) For a reader, perhaps the most useful function of both silent letters and the availability of different spellings for one sound is to distinguish homophones from each other (son/sun /sʌn/, hour/our /ˈaʊə/, rite/write/right/ wright /raɪt/). (2) Silent is the regular way of indicating that the preceding V is “heavy” (car /kɑː/), unless is followed by silent , in which case the previous V is nearly always r-tense (care /keə/). But remember that syllable-initial or is of course not silent (caring /ˈkeərɪŋ/, starring /ˈstɑːrɪŋ/, marrying /ˈmærɪɪŋ/). (3) Initial is silent in some words, especially if a very similar word, with , exists in Spanish (heir /eә/, hour /aʊә/, honour /ˈɒnә/, honest /ˈɒnɪst/) (see § 4.2.2.5). Otherwise, /h/ is usually pronounced (hair /heә/, hat /hæt/, head /hed/, horrid /ˈhɒrɪd/, harmony /ˈhɑːmәni/).

112 Graphotactic constraints also affect vowels. One, for instance, requires to be almost always followed by another letter, as in argue /ˈɑːɡjuː/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of consonants

335

(4) In certain compound words in which the first part (in isolation) ends in a C sound and the second begins with another one, certain letter pairs are pronounced complete (as in timetable /ˈtaɪmˌteɪbl ̩/), while others are not, as in grandson /ˈɡrænsʌn/, where the is pronounced but the is silent. Similarly the before in handkerchief /ˈhæŋkətʃɪf/ is silent. In blackguard /ˈblæɡɑːd/ is silent but is pronounced. (5) C-doubling: one might wish to consider one member of the doublet as pronounced and the other as silent. A double C between Vs marks the previous one as most likely lax. So, compare planning /ˈplænɪŋ/ with planing /ˈpleɪnɪŋ/, is lax in the former word, but tense in the latter. On the other hand, doubling often simply reflects a Latin word, part of whose prefix (sub) became silent, resulting eventually in English spellings like succumb /səˈkʌm/, instead of subcumb.113 (6) Many words begin or end with consonant digraphs (or trigraphs) consisting of a sequence of two (or three) consecutive consonant letters, both of which presumably used to be pronounced but no longer are. Usually one letter has its usual pronunciation and the other is silent and has no practical function. Take for instance the digraph in the word character /kærɪktǝ/: it would be pronounced the same without the letter (which simply reflects Greek origin). In other cases, the digraph has a pronunciation that is not the same as either letter in isolation, as in the case of (sing /sɪŋ/). Cases like this suggest that there are not enough single letters in the alphabet to distinguish each sound in the language. There are also cases of a digraph that is “shared” by two very similar words, one of which is longer and with full digraph pronounced, the other being shorter and with only one digraph part pronounced. This is what occurs to the second in bomb /bɒm/: it is silent in this word because it occurs in word-final position, but it is audible word-medially, as in the verb bombard /bɒmˈbɑːd/, so the connection between the words is more obvious. Similarly, the of autumn is silent /ˈɔːtəm/, but it is pronounced in autumnal /ɔːˈtʌmnəl/. In the following list, silent letters mostly in digraphs and trigraphs are grouped according to their position in the word. You will see that usually, the first or last letter of a word-initial and a word-final digraph are silent, but there also exist other possibilities:

113 A double consonant also indicates that the preceding syllable is probably stressed (ˈoffered vs. reˈferred). When is doubled, it is nearly always /s/, while can be /s/ or /z/.

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

336

Predicting Pronunciation from Spelling (and Vice Versa)

AI 7.3 Silent letters a.

Initial silent letter in word-initial two-consonant clusters: – czar /zɑː/ – gnash /næʃ/, gnaw /nɔː/, gnome /nəʊm/ – knead /niːd/, knee /niː/, knell /nel/, knife /naɪf/, knit /nɪt/, knight /naɪt/, knock /nɒk/, knot /nɒt/, know /nəʊ/, knickers /ˈnɪkəz/ – pneumonia /njuːˈməʊnɪə/, pseudo-psyˈchology /ˌsjuːdəʊ saɪˈkɒlədʒi/ – who /huː/, whom /huːm/, whose /huːz/, whole /həʊl/, whore /hɔː/ – write /raɪt/, wrist /rɪst/, wrong /rɒŋ/ b. Second silent letter in word-initial two-consonant clusters: – chaos /ˈkeɪɒs/, character /ˈkærəktə/, chronic /ˈkrɒnɪk/, echo /ˈekəʊ/ – , ghost /ɡəʊst/, ghastly /ˈɡɑːstli/, khaki /ˈkɑːki/ – , two /tuː/, sword /sɔːd/ – what /ˈwɒt/, where /weə/, when /wen/, why /waɪ/ wharf /wɔːf/ c. Penultimate silent letter in syllable-final and word-final two-consonant clusters: – debt /det/, debtor /ˈdetə/, doubt /daʊt/ – indict /ɪnˈdaɪt/ – align /əˈlaɪn /, reign /reɪn/, phlegm /flem/, diaphragm /ˈdaɪəfræm/ – , chalk /tʃɔːk/, folk /fəʊk/, calf /kɑːf/ – alm /ɑːm/, calm /kɑːm/, almond /ˈɑːmənd /, salmon /ˈsæmən/ – receipt /rɪˈsiːt/ – viscount /ˈvaɪkaʊnt/ – isle /aɪl/, island /ˈaɪlənd/ aisle /aɪl/ d. Last silent letter in word-final two-consonant clusters: – autumn /ˈɔːtəm/, column /ˈkɒləm/, solemn /ˈsɒləm/, damn /dæm/, hymn /hɪm/, condemn /kənˈdem/ – bomb /bɒm/, crumb /krʌm/, climb /klaɪm/, comb /kəʊm/, tomb /tuːm/ e. Two silent letters in syllable-final and word-final two-consonant clusters: – dough /dəʊ /, though /ðəʊ/, through /θruː/, plough /plaʊ/, thorough /ˈθʌrə/, high /haɪ/, daughter /dɔːtǝ/ f. Two pre-final silent letters before word-final consonants: – bright /braɪt/, height /haɪt/, weight /weɪt/, fought /ˈfɔːt/, caught /ˈkɔːt/ g. Other distributions of silent letters: – subtle /ˈsʌtl ̩/ – muscle /ˈmʌsl ̩/ – twelfth /twelfθ/

Brought to you by | Universität Osnabrück Authenticated Download Date | 1/16/16 2:15 PM

Spelling-to-sound correspondences of consonants

337

– champagne /ʃæmˈpeɪn/ – and knight /naɪt/ – (and 2nd ) colonel /ˈkɜːnl ̩/ –